Psychological perspectives on expertise

PSYCHOLOGICAL PERSPECTIVES ON EXPERTISE

Topic EditorsGuillermo Campitelli, Michael H. Connors, Merim Bilalic and David Zachary Hambrick

PSYCHOLOGY

http://journal.frontiersin.org/researchtopic/psychological-perspectives-on-expertise-1885



http://www.frontiersin.org/psychology


Frontiers in Psychology May 2015 | Psychological perspectives on expertise | 1

ABOUT FRONTIERSFrontiers is more than just an open-access publisher of scholarly articles: it is a pioneering approach to the world of academia, radically improving the way scholarly research is managed. The grand vision of Frontiers is a world where all people have an equal opportunity to seek, share and generate knowledge. Frontiers provides immediate and permanent online open access to all its publications, but this alone is not enough to realize our grand goals.

FRONTIERS JOURNAL SERIESThe Frontiers Journal Series is a multi-tier and interdisciplinary set of open-access, online journals, promising a paradigm shift from the current review, selection and dissemination processes in academic publishing. All Frontiers journals are driven by researchers for researchers; therefore, they constitute a service to the scholarly community. At the same time, the Frontiers Journal Series operates on a revo-lutionary invention, the tiered publishing system, initially addressing specific communities of scholars, and gradually climbing up to broader public understanding, thus serving the interests of the lay society, too.

DEDICATION TO QUALITYEach Frontiers article is a landmark of the highest quality, thanks to genuinely collaborative interac-tions between authors and review editors, who include some of the world’s best academicians. Research must be certified by peers before entering a stream of knowledge that may eventually reach the public - and shape society; therefore, Frontiers only applies the most rigorous and unbiased reviews.Frontiers revolutionizes research publishing by freely delivering the most outstanding research, evaluated with no bias from both the academic and social point of view.By applying the most advanced information technologies, Frontiers is catapulting scholarly publishing into a new generation.

WHAT ARE FRONTIERS RESEARCH TOPICS?Frontiers Research Topics are very popular trademarks of the Frontiers Journals Series: they are collections of at least ten articles, all centered on a particular subject. With their unique mix of varied contributions from Original Research to Review Articles, Frontiers Research Topics unify the most influential researchers, the latest key findings and historical advances in a hot research area! Find out more on how to host your own Frontiers Research Topic or contribute to one as an author by contacting the Frontiers Editorial Office: [email protected]

FRONTIERS COPYRIGHT STATEMENT© Copyright 2007-2015 Frontiers Media SA. All rights reserved.

All content included on this site, such as text, graphics, logos, button icons, images, video/audio clips, downloads, data compilations and software, is the property of or is licensed to Frontiers Media SA (“Frontiers”) or its licensees and/or subcontractors. The copyright in the text of individual articles is the property of their respective authors, subject to a license granted to Frontiers.

The compilation of articles constituting this e-book, wherever published, as well as the compilation of all other content on this site, is the exclusive property of Frontiers. For the conditions for downloading and copying of e-books from Frontiers’ website, please see the Terms for Website Use. If purchasing Frontiers e-books from other websites or sources, the conditions of the website concerned apply.

Images and graphics not forming part of user-contributed materials may not be downloaded or copied without permission.

Individual articles may be downloaded and reproduced in accordance with the principles of the CC-BY licence subject to any copyright or other notices. They may not be re-sold as an e-book.

As author or other contributor you grant a CC-BY licence to others to reproduce your articles, including any graphics and third-party materials supplied by you, in accordance with the Conditions for Website Use and subject to any copyright notices which you include in connection with your articles and materials.

All copyright, and all rights therein, are protected by national and international copyright laws.

The above represents a summary only. For the full conditions see the Conditions for Authors and the Conditions for Website Use.

Cover image provided by Ibbl sarl, Lausanne CH

ISSN 1664-8714ISBN 978-2-88919-520-6DOI 10.3389/978-2-88919-520-6



http://www.frontiersin.org/

http://creativecommons.org/licenses/by/4.0/



Topic Editors:Guillermo Campitelli, Edith Cowan University, AustraliaMichael H. Connors, Macquarie University, AustraliaMerim Bilalic, Alpen Adria University Klagenfurt, AustriaDavid Zachary Hambrick, Michigan State University, USA

Experts are persons who are very knowledgeable about or skilful in a particular area. The aim of this Research Topic is to advance knowledge in the understanding of the phenomenon of expertise by putting together different lines of research that directly or indirectly study expertise.

Herbert Simon’s expertise studies initiated two lines of research. One is interested in elucidating the cognitive processes underlying expertise, and the other investigates how expertise develops. These lines of research started with studies comparing experts and novices in chess, and then they extended to numerous areas of expertise such as music, medical diagnosis, sports, arts and sciences.

In the field of judgment and decision making researchers investigate the quality of judgments and decisions of experts in different professions (e.g., clinical psychologists, medical practitioners, judges, meteorologists, stock brokers).

Those lines of research explicitly investigate the topic of expertise, but there are other research areas that make a substantial contribution to understanding expertise. Scholars in language acquisition and in face perception, for example, investigate cognitive processes and development of expertise in areas in which almost everyone becomes an expert. Furthermore, skill acquisition research informs in detail about short term cognitive changes that may be important to understand how expertise develops.

We are interested in original research that advances knowledge in the understanding of decision making, cognitive processes and development of expertise in sports, intellectual games, arts, scientific disciplines and professions, as well as expertise in cognitive abilities such as perception, memory, attention, language and imagery.

PSYCHOLOGICAL PERSPECTIVES ON EXPERTISE




We are also interested in theoretical articles in any of these areas, articles that describe computational or mathematical models of expertise, and articles offering a framework that would guide expertise research. Articles that offer integrative approaches of some of the areas described above are strongly encouraged.

The goal of this Research Topic is to produce a hallmark piece of work in the field of expertise, which complements and does not overlap with the “Neural implementations of expertise” Research Topic in Frontiers in Human Neuroscience.




Table of Contents

07 Psychological Perspectives on ExpertiseGuillermo Campitelli, Michael H. Connors, Merim Bilalić and David Z. Hambrick

11 Expert vs. Novice Differences in the Detection of Relevant Information During a Chess Game: Evidence From Eye MovementsHeather Sheridan and Eyal M. Reingold

17 Expert and Competent Non-Expert Visual Cues During Simulated Diagnosis in Intensive CareClare McCormack, Mark W. Wiggins, Thomas Loveday and Marino Festa

24 Spontaneously Spotting and Applying Shortcuts in Arithmetic—A Primary School Perspective on ExpertiseClaudia Godau, Hilde Haider, Sonja Hansen, Torsten Schubert, Peter A. Frensch and Robert Gaschler

35 The Einstellung Effect in Anagram Problem Solving: Evidence From Eye MovementsJessica J. Ellis and Eyal M. Reingold

42 Interference between Face and Non-Face Domains of Perceptual Expertise: A Replication and ExtensionKim M. Curby and Isabel Gauthier

53 Visual Appearance Interacts with Conceptual Knowledge in Object RecognitionOlivia S. Cheung and Isabel Gauthier

64 Segmentation of Dance Movement: Effects of Expertise, Visual Familiarity, Motor Experience and MusicBettina E. Bläsing

75 Expertise Affects Representation Structure and Categorical Activation of Grasp Postures in ClimbingBettina E. Bläsing, Iris Güldenpenning, Dirk Koester and Thomas Schack

86 Timing Skills and Expertise: Discrete and Continuous Timed Movements Among Musicians and AthletesThenille Braun Janzen, William Forde Thompson, Paolo Ammirante and Ronald Ranvaud

97 Trait-Based Cue Utilization and Initial Skill Acquisition: Implications for Models of the Progression to ExpertiseMark W. Wiggins, Sue Brouwers, Joel Davies and Thomas Loveday

105 Understanding Expertise and Non-Analytic Cognition in Fingerprint Discriminations Made by HumansMatthew B. Thompson, Jason M. Tangen and Rachel A. Searston




108 Expert Analogy Use in a Naturalistic SettingDonald R. Kretz and Daniel C. Krawczyk

116 Explaining the Abundance of Distant Analogies in Naturalistic Observations of ExpertsMáximo Trench

118 Can Taking the Perspective of an Expert Debias Human Decisions? The Case of Risky and Delayed GainsMichał Białek and Przemysław Sawicki

126 The Geometry of ExpertiseMaría J. Leone, Diego Fernandez Slezak, Guillermo A. Cecchi and Mariano Sigman

135 Expertise and the Representation of SpaceMichael H. Connors and Guillermo Campitelli

137 Playing Off the Curve-Testing Quantitative Predictions of Skill Acquisition Theories in Development of Chess PerformanceRobert Gaschler, Johanna Progscha, Kieran Smallbone, Nilam Ram and Merim Bilalić

148 Checkmate to Deliberate Practice: The Case of Magnus CarlsenFernand Gobet and Morgan H. Ereku

151 The Influence of Deliberate Practice on Musical Achievement: A Meta-AnalysisFriedrich Platz, Reinhard Kopiez, Andreas C. Lehmann and Anna Wolf

164 Facing Facts About Deliberate PracticeDavid Z. Hambrick, Erik M. Altmann, Frederick L. Oswald, Elizabeth J. Meinz and Fernand Gobet

166 Training Principles to Advance ExpertiseAlice F. Healy, James A. Kole and Lyle E. Bourne Jr

170 The Acquisition of Expertise in the Classroom: Are Current Models of Education Appropriate?Craig P. Speelman

173 Designing a “Better” Brain: Insights From Experts and SavantsFernand Gobet, Allan Snyder, Terry Bossomaier and Mike Harré

176 2011 Space Odyssey: Spatialization as a Mechanism to Code Order Allows a Close Encounter Between Memory Expertise and Classic Immediate Memory StudiesAlessandro Guida and Magali Lavielle-Guida

181 Teachers’ Expertise in Feedback Application Adapted to the Phases of the Learning ProcessEva Christophel, Robert Gaschler and Wolfgang Schnotz

185 “No Level Up!”: No Effects of Video Game Specialization and Expertise on Cognitive PerformanceFernand Gobet, Stephen J. Johnston, Gabriella Ferrufino, Matthew Johnston, Michael B. Jones, Antonia Molyneux, Argyrios Terzis and Luke Weeden

194 Multi-Domain Computerized Cognitive Training Program Improves Performance of Bookkeeping Tasks: A Matched-Sampling Active-Controlled TrialAmit Lampit, Claus Ebster and Michael Valenzuela

201 On the Effect of Chess Training on Scholastic AchievementWilliam M. Bart




204 Restricting Range Restricts ConclusionsNemanja Vaci, Bartosz Gula and Merim Bilalić

208 A Proposed Integration of the Expert Performance and Individual Differences Approaches to the Study of Elite PerformanceScott Barry Kaufman

211 Expertise: Defined, Described, ExplainedLyle E. Bourne Jr., James A. Kole and Alice F. Healy

214 Studying Real-World Perceptual ExpertiseJianhong Shen, Michael L. Mack and Thomas J. Palmeri

220 Metacognition and Action: A New Pathway to Understanding Social and Cognitive Aspects of Expertise in SportTadhg E. MacIntyre, Eric R. Igou, Mark J. Campbell, Aidan P. Moran and James Matthews

232 In Praise of Conscious Awareness: A New Framework for the Investigation of “Continuous Improvement” in Expert AthletesJohn Toner and Aidan Moran

237 An Adaptive Toolbox Approach to the Route to Expertise in SportRita F. de Oliveira, Babett H. Lobinger and Markus Raab



EDITORIALpublished: 10 March 2015

doi: 10.3389/fpsyg.2015.00258

Frontiers in Psychology | www.frontiersin.org March 2015 | Volume 6 | Article 258

Edited and reviewed by:

Bernhard Hommel,

Leiden University, Netherlands

*Correspondence:

Guillermo Campitelli,

[email protected]

Specialty section:

This article was submitted to

Cognition, a section of the journal

Frontiers in Psychology

Received: 20 February 2015

Accepted: 21 February 2015

Published: 10 March 2015

Citation:

Campitelli G, Connors MH, Bilalic M

and Hambrick DZ (2015)

Psychological perspectives on

expertise. Front. Psychol. 6:258.

doi: 10.3389/fpsyg.2015.00258

Psychological perspectives onexpertise

Guillermo Campitelli 1*, Michael H. Connors 2, 3, Merim Bilalic 4 and David Z. Hambrick 5

1 School of Psychology and Social Science, Edith Cowan University, Joondalup, WA, Australia, 2Department of Cognitive

Science, ARC Centre of Excellence in Cognition and its Disorders, Macquarie University, Sydney, NSW, Australia, 3Dementia

Collaborative Research Centre, School of Psychiatry, University of New South Wales, Sydney, NSW, Australia, 4Department

of General Psychology and Cognitive Science, Institute of Psychology, Alpen Adria University Klagenfurt, Klagenfurt, Austria,5Department of Psychology, Michigan State University, East Lansing, MI, USA

Keywords: expertise, expert performance, expert cognitive processes, skill acquisition, skill transfer, training,

deliberate practice

Introduction

This Research Topic sought to advance psychological understanding of expertise by drawingtogether lines of research from many different domains of expertise. The outcome is a collectionof 35 articles in such diverse areas as chess, music, perception, teaching, intensive-care diagnosis,video-games, sports, dance, mathematics, climbing, and fingerprint analysis.

The articles can be classed into five broad categories based on their focus: (a) the cognitive pro-cesses in expertise, (b) the development of expertise, (c) the relationship between expertise andgeneral cognition, (d) the transfer of skills between domains, and (e) methodological issues andframeworks in expertise research. We give a brief overview of the research across these five themes.

Cognitive Processes in Expertise

Articles in this research topic used a number of different methodologies to investigate cognitiveprocesses in expertise. Four articles examined experts’ eye movements as a way of studying whatexperts focus on when performing domain-relevant tasks. Sheridan and Reingold (2014), for exam-ple, found that expert chess players rapidly differentiate regions of the board that are relevant to thebest move from irrelevant ones. Similarly,McCormack et al. (2014) found that expert intensive-carephysicians directed their attention to more relevant areas of the situation, compared to competentnon-experts. In addition, Godau et al. (2014) found that experts in arithmetic problem solvingspontaneously used arithmetic shortcuts. Finally, Ellis and Reingold (2014) examined the Ein-stellung effect (i.e., where the first idea that comes to mind blocks finding the best solution to aproblem) using this methodology and noted its relevance to understanding expert flexibility (seeBilalic and McLeod, 2014).

Two articles focused on perceptual expertise. Curby and Gauthier (2014) found that acquiringexpertise with a category of stimuli (i.e., car expertise) increases the interference between the visualprocessing of other familiar stimuli (e.g., faces) and that of the learned category (cars). In a studywith novel objects, Cheung and Gauthier (2014) found that acquiring perceptual expertise involvesintegrating perceptual and conceptual representations of stimuli.

Four studies investigated expertise involving physical movements. In a study on dancing,Bläsing (2015) found that making a sequence of movements influenced the subsequent percep-tion of that sequence, but not to the same degree if one was a dancing expert. In a study ofclimbing experts, Bläsing et al. (2014) found that expertise was associated with better percep-tion of climbing holds and action-relevant objects. In a study of athletes and musicians, BraunJanzen et al. (2014) showed that training affects performance involving timing and rhythmic

| 7

http://www.frontiersin.org/Psychology

http://www.frontiersin.org/Psychology/editorialboard






http://www.frontiersin.org

http://www.frontiersin.org/Psychology/archive

https://creativecommons.org/licenses/by/4.0/

mailto:[email protected]

http://dx.doi.org/10.3389/fpsyg.2015.00258

http://www.frontiersin.org/journal/10.3389/fpsyg.2015.00258/abstract

http://community.frontiersin.org/people/u/31638




Campitelli et al. Psychological perspectives on expertise

movements: athletes were more precise making continuousmovements, whereas musicians were more precise for discretemovements. In a study of skill acquisition in a flight simulator,Wiggins et al. (2014) found that a general capacity for acquiringand using cues was related to performance in landing an aircraftin the simulator.

Finally, four articles focused on experts’ pattern recognition(the ability to identify meaningful relationships in complex stim-uli). In a review of research on fingerprint experts, Thompsonet al. (2014) concluded that such expertise relies on rapid patternrecognition and discrimination rather than in analytic thinking.In an observational study, Kretz and Krawczyk (2014) found thatacademic economists use many analogies in research meetings.Trench (2014), however, suggests that these results may be dueto the naturalistic setting of the study, rather than expertise perse. Bialek and Sawicki (2014) showed that participants asked totake an expert perspective become more risk aversive and patientin decision making tasks. Finally, Leone et al. (2014) examinedthe relationship between expertise and representations of spaceusing a large dataset of chess games from an internet server. Theyfound that novices, relative to experts, use strategies to reducetheir cognitive load (see Connors and Campitelli, 2014, for acommentary).

Development of Expertise

Six studies examined how expertise is developed. Gaschler et al.(2014) examined the learning curves in skill acquisition by ana-

lyzing the tournament performance of 1383 chess players over 10years. They found that exponential learning curves better fittedplayers’ improvements over time than power function learningcurves. Gobet and Ereku (2014) discussed the case of MagnusCarlsen, current world chess champion, and argue that his levelof performance cannot be accounted for by the deliberate prac-tice account, which suggests that amount of deliberate practice isthe critical determinant of expertise.

Citing limitations in an earlier meta-analysis by Hambricket al. (2014a), Platz et al. (2014) conducted a meta-analysis onthe influence of deliberate practice in musical achievement. Theyfound amoderate average effect size (rc = 0.61), which they inter-pret as showing the importance of deliberate practice. In responseto Platz et al.’s criticisms, Hambrick et al. (2014b) noted a num-ber of conceptual problems in Platz et al.’s arguments and observehow Platz et al.’s findings can also be interpreted to show thatpractice, while undoubtedly an important factor in expertise, isnot the sole determinant.Healy et al. (2014) proposed a number of training principles for

developing expertise. These include the acquisition of expertise(e.g., scheduling of feedback), retention of expertise (e.g., itemchunking, depth of processing) and transfer (e.g., variability ofpractice, seeding the knowledge base). Finally, Speelman (2014)argued that treating numeracy as a form of expertise and usingcomputer programs in teaching would address some shortcom-ings in current teaching and, in particular, foster a greater focuson practice and feedback in learning.

Expertise and General Cognition

Three articles examined the relationship between expertise andgeneral cognition. First, Gobet et al. (2014b) discussed howartificial intelligence and engineering could be used to designa brain. Based on expertise research, they propose that a bet-ter brain would have less concepts and more low-level per-ceptual processing. Second, Guida and Lavielle-Guida (2014)combined findings from memory research with the normalpopulation with theories of expert memory. They argue thata less sophisticated version of the spatial method of lociused by memory experts is also used by ordinary people toencode items in working memory. Third, Christophel et al.(2014) observed that amount of teaching experience is a verypoor predictor of a teacher’s actual effectiveness, including, forexample, the teacher’s ability to offer constructive feedback tostudents.

Transfer of Skills

Three articles examined the transfer of skills across domains.First, Gobet et al. (2014a) investigated the possibility oftraining transfer from videogame playing to selective atten-tion and working memory capacity. Consistent with overa century of research, there was no evidence for transfer,even in videogame experts. In contrast, Lampit et al. (2014)reported evidence for transfer of computerized cognitive train-ing to a bookkeeping task. Finally, Bart (2014) reviewedresearch published after Gobet and Campitelli’s (2006) crit-ical review on the effects of chess education, showing sta-tistically significant effects of chess education on academicachievement.

Frameworks, Recommendations, and

Methodology

Seven articles discussed expertise in general. First, Vaci et al.(2014) consider alternative approaches to studying expertise,and in particular, how studying only individuals from highlyrestricted ranges of skill may yield different findings thanstudying individuals who represent wider ranges of skill. Sec-ond, Kaufman (2014) identifies points of disagreement andagreement in different views of expertise and suggests somedirections for future research. Third, Bourne et al. (2014),categorize expertise as elite, peak, or exceptionally high lev-els of performance on a particular task or within a givendomain.

Fourth, Shen et al. (2014) use birdwatching as an illus-trative example to discuss such issues as selecting an appro-priate domain of perceptual expertise for study, recruitingexperts, assessing their level of expertise, and experimentallytesting the experts’ performance. Fifth, MacIntyre et al. (2014)propose that athletes are not just experts in movement exe-cution but also in planning, metacognition, and reflection.Similarly, Toner and Moran (2014), extending Sutton et al.’s

Frontiers in Psychology | www.frontiersin.org March 2015 | Volume 6 | Article 258 | 8





(2011) framework, argue that expert athletes do not com-pletely automatize their skills and that an important com-ponent of their expertise is to be able to rapidly reflect ontheir movements. Finally, de Oliveira et al. (2014) build uponGigerenzer’s (e.g., Gigerenzer and Goldstein, 1999) heuristic-based approach to decision making. They propose that expertathletes develop a toolbox of heuristics to guide their decisionmaking.

Conclusion

The diversity of articles in this research topic illustratesthe many different approaches to studying expertise. It alsoindicates the keen interest in the topic. We believe thatmany articles in this research topic are of lasting impor-tance and can help to guide future research in the field ofexpertise.

References

Bart, W. M. (2014). On the effect of chess training on scholastic achievement.

Front. Psychol. 5:762. doi: 10.3389/fpsyg.2014.00762

Bialek, M., and Sawicki, P. (2014). Can taking the perspective of an expert debias

human decisions? The case of risky and delayed gains. Front. Psychol. 5:989.

doi: 10.3389/fpsyg.2014.00989

Bilalic, M., and McLeod, P. (2014). Why good thoughts block better ones. Sci. Am.

310, 74–79. doi: 10.1038/scientificamerican0314-74

Bläsing, B. E. (2015). Segmentation of dance movement: effects of expertise,

visual familiarity, motor experience and music. Front. Psychol. 5:1500. doi:

10.3389/fpsyg.2014.01500

Bläsing, B. E., Güldenpenning, I., Koester, D., and Schack, T. (2014). Expertise

affects representation structure and categorical activation of grasp postures in

climbing. Front. Psychol. 5:1008. doi: 10.3389/fpsyg.2014.01008

Bourne, L. E. Jr., Kole, J. A., and Healy, A. F. (2014). Expertise: defined, described,

explained. Front. Psychol. 5:186. doi: 10.3389/fpsyg.2014.00186

Braun Janzen, T., Thompson, W. F., Ammirante, P., and Ranvaud, R. (2014).

Timing skills and expertise: discrete and continuous timed movements

among musicians and athletes. Front. Psychol. 5:1482. doi: 10.3389/fpsyg.2014.

01482

Cheung, O. S., and Gauthier, I. (2014). Visual appearance interacts with

conceptual knowledge in object recognition. Front. Psychol. 5:793. doi:

10.3389/fpsyg.2014.00793

Christophel, E., Gaschler, R., and Schnotz, W. (2014). Teachers’ expertise in feed-

back application adapted to the phases of the learning process. Front. Psychol.

5:858. doi: 10.3389/fpsyg.2014.00858

Connors, M. H., and Campitelli, G. (2014). Expertise and the representation of

space. Front. Psychol. 5:270. doi: 10.3389/fpsyg.2014.00270

Curby, K. M., and Gauthier, I. (2014). Interference between face and non-face

domains of perceptual expertise: a replication and extension. Front. Psychol.

5:955. doi: 10.3389/fpsyg.2014.00955

de Oliveira, R. F., Lobinger, B. H., and Raab, M. (2014). An adaptive tool-

box approach to the route to expertise in sport. Front. Psychol. 5:709. doi:

10.3389/fpsyg.2014.00709

Ellis, J. J., and Reingold, E. M. (2014). The Einstellung effect in anagram

problem solving: evidence from eye movements. Front. Psychol. 5:679. doi:

10.3389/fpsyg.2014.00679

Gaschler, R., Progscha, J., Smallbone, K., Ram, N., and Bilalic, M. (2014).

Playing off the curve - testing quantitative predictions of skill acquisition

theories in development of chess performance. Front. Psychol. 5:923. doi:

10.3389/fpsyg.2014.00923

Gigerenzer, G., and Goldstein, D. G. (1999). “Betting on one good reason: the take

the best heuristic,” in Simple Heuristics that Make us Smart, eds G. Gigerenzer,

P. M. Todd, and The ABCResearch Group (New York, NY: Oxford University),

75–95.

Gobet, F., and Campitelli, G. (2006). “Education and chess: a critical review,” in

Chess and Education: Selected Essays from the Koltanowski Conference, ed T.

Redman (Dallas, TX: Chess Program at the University of Texas at Dallas),

124–143.

Gobet, F., and Ereku, M. H. (2014). Checkmate to deliberate practice: the case of

Magnus Carlsen. Front. Psychol. 5:878. doi: 10.3389/fpsyg.2014.00878

Gobet, F., Johnston, S. J., Ferrufino, G., Johnston, M., Jones, M. B., Molyneux,

A., et al. (2014a). “No level up!”: no effects of video game specialization

and expertise on cognitive performance. Front. Psychol. 5:1337. doi:

10.3389/fpsyg.2014.01337

Gobet, F., Snyder, A., Bossomaier, T., and Harré, M. (2014b). Designing a “bet-

ter” brain: insights from experts and savants. Front. Psychol. 5:470. doi:

10.3389/fpsyg.2014.00470

Godau, C., Haider, H., Hansen, S., Schubert, T., Frensch, P. A., and Gaschler,

R. (2014). Spontaneously spotting and applying shortcuts in arithmetic—

a primary school perspective on expertise. Front. Psychol. 5:556. doi:

10.3389/fpsyg.2014.00556

Guida, A., and Lavielle-Guida, M. (2014). 2011 space odyssey: spatialization

as a mechanism to code order allows a close encounter between memory

expertise and classic immediate memory studies. Front. Psychol. 5:573. doi:

10.3389/fpsyg.2014.00573

Hambrick, D. Z., Altmann, E. M., Oswald, F. L., Meinz, E. J., and Gobet, F.

(2014b). Facing facts about deliberate practice. Front. Psychol. 5:751. doi:

10.3389/fpsyg.2014.00751

Hambrick, D. Z., Oswald, F. L., Altmann, E. M., Meinz, E. J., Gobet, F., and

Campitelli, G. (2014a). Deliberate practice: is that all it takes to become an

expert? Intelligence 45, 34–45. doi: 10.1016/j.intell.2013.04.001

Healy, A. F., Kole, J. A., and Bourne, L. E. Jr. (2014). Training principles to advance

expertise. Front. Psychol. 5:131. doi: 10.3389/fpsyg.2014.00131

Kaufman, S. B. (2014). A proposed integration of the expert performance and indi-

vidual differences approaches to the study of elite performance. Front. Psychol.

5:707. doi: 10.3389/fpsyg.2014.00707

Kretz, D. R., and Krawczyk, D. C. (2014). Expert analogy use in a naturalistic

setting. Front. Psychol. 5:1333. doi: 10.3389/fpsyg.2014.01333

Lampit, A., Ebster, C., and Valenzuela, M. (2014). Multi-domain comput-

erized cognitive training program improves performance of bookkeeping

tasks: a matched-sampling active-controlled trial. Front. Psychol. 5:794. doi:

10.3389/fpsyg.2014.00794

Leone, M. J., Fernandez Slezak, D., Cecchi, G. A., and Sigman, M. (2014). The

geometry of expertise. Front. Psychol. 5:47. doi: 10.3389/fpsyg.2014.00047

MacIntyre, T. E., Igou, E. R., Campbell, M. J., Moran, A. P., and Matthews, J.

(2014). Metacognition and action: a new pathway to understanding social

and cognitive aspects of expertise in sport. Front. Psychol. 5:1155. doi:

10.3389/fpsyg.2014.01155

McCormack, C., Wiggins, M. W., Loveday, T., and Festa, M. (2014). Expert and

competent non-expert visual cues during simulated diagnosis in intensive care.


Platz, F., Kopiez, R., Lehmann, A. C., and Wolf, A. (2014). The influence of delib-

erate practice on musical achievement: a meta-analysis. Front. Psychol. 5:646.

doi: 10.3389/fpsyg.2014.00646

Shen, J., Mack, M. L., and Palmeri, T. J. (2014). Studying real-world perceptual

expertise. Front. Psychol. 5:857. doi: 10.3389/fpsyg.2014.00857

Sheridan, H., and Reingold, E. M. (2014). Expert vs. novice differences in the

detection of relevant information during a chess game: evidence from eye

movements. Front. Psychol. 5:941. doi: 10.3389/fpsyg.2014.00941

Speelman, C. P. (2014). The acquisition of expertise in the classroom:

are current models of education appropriate? Front. Psychol. 5:580. doi:

10.3389/fpsyg.2014.00580

Sutton, J., Mcilwain, D., Christensen, W., and Geeves, A. (2011). Applying

intelligence to the reflexes: embodied skills and habits between dreyfus and

descartes. J. Br. Soc. Phenomenol. 42, 78–103. doi: 10.1080/00071773.2011.

11006732






Thompson, M. B., Tangen, J. M., and Searston, R. A. (2014). Understanding exper-

tise and non-analytic cognition in fingerprint discriminationsmade by humans.


Toner, J., andMoran, A. (2014). In praise of conscious awareness: a new framework

for the investigation of “continuous improvement” in expert athletes. Front.

Psychol. 5:769. doi: 10.3389/fpsyg.2014.00769

Trench, M. (2014). Explaining the abundance of distant analogies in natural-

istic observations of experts. Front. Psychol. 5:1487. doi: 10.3389/fpsyg.2014.

01487

Vaci, N., Gula, B., and Bilalic, M. (2014). Restricting range restricts conclusions.


Wiggins, M. W., Brouwers, S., Davies, J., and Loveday, T. (2014). Trait-

based cue utilization and initial skill acquisition: implications for models

of the progression to expertise. Front. Psychol. 5:541. doi: 10.3389/fpsyg.

2014.00541

Conflict of Interest Statement: The authors declare that the research was con-

ducted in the absence of any commercial or financial relationships that could be

construed as a potential conflict of interest.

Copyright © 2015 Campitelli, Connors, Bilalic and Hambrick. This is an open-access

article distributed under the terms of the Creative Commons Attribution License (CC

BY). The use, distribution or reproduction in other forums is permitted, provided the

original author(s) or licensor are credited and that the original publication in this

journal is cited, in accordance with accepted academic practice. No use, distribution

or reproduction is permitted which does not comply with these terms.










ORIGINAL RESEARCH ARTICLEpublished: 25 August 2014

doi: 10.3389/fpsyg.2014.00941

Expert vs. novice differences in the detection of relevantinformation during a chess game: evidence from eyemovementsHeather Sheridan1* and Eyal M. Reingold2

1 School of Psychology, University of Southampton, UK2

Edited by:

Merim Bilalic, Alpen Adria UniversityKlagenfurt, Austria

Reviewed by:

Ramesh Kumar Mishra, Universityof Hyderabad, IndiaJochen Musch, University ofDuesseldorf, GermanyPeter McLeod, Oxford University,UK

*Correspondence:

Heather Sheridan, School ofPsychology, University ofSouthampton, Building 44, HighfieldCampus, Southampton, SO17 1BJ,UKe-mail: [email protected]

The present study explored the ability of expert and novice chess players to rapidlydistinguish between regions of a chessboard that were relevant to the best move on theboard, and regions of the board that were irrelevant. Accordingly, we monitored the eyemovements of expert and novice chess players, while they selected white’s best movefor a variety of chess problems. To manipulate relevancy, we constructed two differentversions of each chess problem in the experiment, and we counterbalanced these versionsacross participants. These two versions of each problem were identical except that asingle piece was changed from a bishop to a knight. This subtle change reversed therelevancy map of the board, such that regions that were relevant in one version of theboard were now irrelevant (and vice versa). Using this paradigm, we demonstrated thatboth the experts and novices spent more time fixating the relevant relative to the irrelevantregions of the board. However, the experts were faster at detecting relevant informationthan the novices, as shown by the finding that experts (but not novices) were able todistinguish between relevant and irrelevant information during the early part of the trial.These findings further demonstrate the domain-related perceptual processing advantageof chess experts, using an experimental paradigm that allowed us to manipulate relevancyunder tightly controlled conditions.

Keywords: visual expertise, expert performance, chess, eye movements, attention, relevancy

The remarkable perceptual skill of experts is exemplified by expertradiologists who can detect abnormalities in chest X-rays thatwere briefly presented for only 200 ms (Kundel and Nodine,1975), and by chess experts who can memorize chessboardsthat were presented for only a few seconds (De Groot, 1946,1965; Chase and Simon, 1973a,b). Furthermore, while examiningvisual displays that require multiple eye fixations for encoding,experts are adept at rapidly focusing their attention on relevantareas, such that radiologists can rapidly fixate on abnormalities(Kundel et al., 2008), and chess experts can rapidly fixate on thebest move on a chessboard (Charness et al., 2001). Given thatthis fast extraction of relevant information is a key componentof skilled performance in many different domains of expertise(for a review see Reingold and Sheridan, 2011), the goal of thepresent experiment is to further explore expert/novice differencesin the detection of relevant information during a chess game.Accordingly, we will begin by briefly reviewing prior work on per-ceptual skill in the domain of chess, and we will then describe thepresent study’s paradigm and rationale.

Of relevance to the present study, there is a long history ofstudying the perceptual component of expertise in the domainof chess (for reviews, see Charness, 1992; De Groot and Gobet,1996; Reingold and Charness, 2005; Gobet and Charness, 2006;Reingold and Sheridan, 2011). The chess domain providesnumerous methodological advantages, such as well-segmented

visual stimuli for eye movement interest areas, and an officialrating system for objectively quantifying levels of expertise (Elo,1965, 1986). Capitalizing on these methodological advantages,chess expertise has been linked to numerous perceptual pro-cessing advantages, including superior memory performance forbriefly presented chessboards (De Groot, 1946, 1965; Chase andSimon, 1973a,b), the ability to process chess configurations auto-matically and in parallel (Reingold et al., 2001b), and a largervisual span such that experts process larger configurations ofpieces than novices (Reingold et al., 2001a). Consistent with thesebehavioral and eye movement findings, neuroimaging work hasshown expert/novice differences in brain activation in regionsassociated with object and pattern recognition (Bilalic et al.,2010a, 2011a,b, 2012). Taken together, the above findings col-lectively support the view that perceptual skill is a key aspectof expertise in chess (De Groot, 1946, 1965; Chase and Simon,1973a,b) as well as in other domains of visual expertise (for areview see Reingold and Sheridan, 2011).

To provide a theoretical account of the perceptual skill shownby chess experts, Chase and Simon (1973a,b) proposed thatthrough thousands of hours of practice chess experts acquirememories for a large number of “chunks,” which consist ofgroups of chess pieces, and these chunks are supplemented bylarger memory structures called templates (Gobet and Simon,1996, 2000). Such memory structures facilitate performance by

www.frontiersin.org August 2014 | Volume 5 | Article 941 |

Department of Psychology, University of Toronto at Mississauga, Mississauga, ON, Canada

11




http://www.frontiersin.org/Psychology/about







http://www.frontiersin.org/Cognition/archive

Sheridan and Reingold Relevancy detection in chess

enabling chess players to rapidly retrieve useful information,such as advantageous strategies and moves. Thus, Chase andSimon (1973a,b) argued that chess experts use their memoryfor chess-configurations to constrain their search for a move tothe most promising candidates, rather than performing a slowand exhaustive search of all of the possible moves on the board.This theoretical perspective echoes earlier arguments by De Groot(1946, 1965) that chess expertise stems from advantages in mem-ory and perception, rather than from a greater breadth and depthof search during problem solving.

Consistent with this hypothesis that chess experts rely on theirmemory for chess configurations to efficiently guide their searchfor the best move, the eye movements of chess experts revealthat they can rapidly fixate on information that is relevant to thebest move on the board (Tikhomirov and Poznyanskaya, 1966;Simon and Barenfeld, 1969; Charness et al., 2001; Reingold andCharness, 2005). For example, to examine the impact of relevancyon eye movements, Charness et al. (2001) monitored the eyemovements of expert players (mean Elo rating = 2238) and inter-mediate players (mean Elo = 1786) while they selected white’sbest move in a series of chess problems (henceforth, the choose-a-move task). Compared to intermediate players, the expertsproduced a greater proportion of fixations on pieces that were rel-evant to the best move, and this advantage of the experts emergedas early as the first five fixations in the trial (for a discussionof similar findings, Tikhomirov and Poznyanskaya, 1966; Simonand Barenfeld, 1969). As a follow-up to Charness et al. (2001),Reingold and Charness (2005) analyzed the first 10 s of choose-a-move trials to demonstrate that experts rapidly completed aperceptual encoding phase (characterized by shorter fixations)and then shifted to a subsequent problem-solving stage (charac-terized by longer fixations). In marked contrast, the intermediatescontinued to display shorter fixations throughout the 10-s period,which indicates that they needed more time to complete the per-ceptual encoding phase. Taken together, these studies indicate thatchess experts are more efficient at encoding chess configurationsduring a choose-a-move task.

Beyond the choose-a-move task, there is further evidence thatexperts are better than novices at rapidly encoding relevant chessconfigurations. For example, in a memorization task, De Grootand Gobet (1996) demonstrated that the number and total dura-tion of fixations on chess pieces was at least partially correlatedwith the relevance of these pieces to the position, and the magni-tude of this correlation increased as a function of skill. Moreover,using a chess-related visual search task that required participantsto search for relevant pieces on a chessboard, Bilalic et al. (2010a)revealed that chess experts (but not novices) were able to rapidlyand exclusively fixate on task-relevant rather than irrelevant fea-tures. Finally, Bilalic et al. (2012) examined relevancy effects ina threat detection task, in which experts and novices had toexamine chessboards to determine the number of black piecesthat were attacking white pieces. The experts displayed a higherpercentage of fixations on relevant objects (i.e., the pieces thatformed a threat relationship) relative to novices, and this differ-ence between experts and novices emerged as early as the firstthree seconds in the trial. Based on these results, Bilalic et al.(2012) concluded that the “experts’ advantage lies in the ability

to immediately focus on relevant objects and relations betweenthem in the environment and ignore the irrelevant ones.”

Building on this prior work, the present study introduces anew paradigm for studying relevancy effects in chess. Similar toprior work (Charness et al., 2001; Reingold and Charness, 2005;Bilalic et al., 2008a,b, 2010b; Sheridan and Reingold, 2013), thepresent paradigm monitored the eye movements of chess playersduring a choose-a-move-task, which is an ecologically valid taskgiven that it resembles the challenges confronting chess playersduring a chess match. To provide a well-controlled manipulationof relevancy, the present paradigm employs two counterbalancedversions of each chess problem, which differed by a single piecesuch that a bishop was changed to a knight (see Appendix Ain Supplementary material for examples). This subtle changereversed the relevancy map of the board, such that the regions thatwere relevant to the best move in one version are irrelevant in theother version, and vice versa. Thus, the present paradigm extendsprior work by employing large relevant and irrelevant regionsof interest that were well-matched on a variety of characteristics(e.g., number of squares, location, etc.).

Our rationale for using this paradigm was to contrast thetime-course of relevancy effects in both novice and expert play-ers. Thus, we asked strong expert players (average Elo = 2223)and novices (unrated club players) to solve a series of problemsthat were designed to be simple enough that both the novice andexpert players could frequently detect the best move on the board.In light of past findings concerning the perceptual encodingadvantage of experts, we expected that the expert’s eye move-ments would reveal an earlier differentiation between relevantand irrelevant regions. Such a finding would provide additionalsupport for the importance of perceptual processing in chess skill,using a new paradigm that afforded a number of methodologicaladvantages.

METHODSPARTICIPANTSForty-one chess players (17 experts and 24 novices) were recruitedfrom online chess forums and from local chess clubs in Torontoand Mississauga (Canada). The mean age was 30 (SD = 14.2)in the expert group, and 27 (SD = 10.0) in the novice group.There was one female player in the expert group, and there werefive female players in the novice group. For the expert players,the average CFC (Canadian Chess Federation) rating was 2223(range = 1876–2580). All of the novice players were unrated clubplayers who were familiar with the rules of chess, but had neverparticipated in a rated chess tournament. All of the participantshad normal or corrected-to-normal vision.

MATERIALS AND DESIGNThere were eight experimental problems (See Appendix A inSupplementary material for the complete list of problems). Tomanipulate relevancy, we constructed two versions of each of theproblems, and these two versions were identical except that asingle piece was changed from a bishop to a knight. As shownin Appendix A in Supplementary material, this subtle changereversed the relevancy map of the board, such that regions thatwere relevant to the best move in one version were no longer

Frontiers in Psychology | Cognition August 2014 | Volume 5 | Article 941 | 12

http://www.frontiersin.org/Cognition




relevant (and vice versa). Similar to Charness et al. (2001), rele-vancy was determined by asking an international master who didnot participate in the study to classify the squares on the boardas either relevant or irrelevant (see also De Groot and Gobet,1996). For example, in the first version of Problem #3 (see panel3a in Appendix A in Supplementary material), the best move onthe board is “Rook to a8 (checkmate),” and the squares associ-ated with this move are located on the left side of the board (seerelevant region in orange), whereas the other side of the boardcontains squares that are irrelevant to the best move (see irrel-evant region in blue). In contrast, in the second version of theproblem (see panel 3b in Appendix A in Supplementary material)we changed a single piece from a bishop to a knight (see pieceinside the dotted lines), such that the best move on the boardbecame “Knight to f4,” and the relevant and irrelevant regionswere reversed. The relevant and irrelevant regions were alwayslocated near the edge of the board, and never overlapped withthe center of the board. The two versions of the problems werecounterbalanced such that each player only saw one version of agiven problem. The same order of problems was used for all play-ers, and each chess player completed a total of eight experimentalproblems. It was always white’s turn to move, and the problemsincorporated a variety of solutions that ranged from checkmateto material gains to defensive tactics.

APPARATUS AND PROCEDUREEye movements were measured with an SR Research EyeLink1000 system with high spatial resolution and a sampling rateof 1000 Hz. The experiment was programmed and analyzedusing SR Research Experiment Builder and Data Viewer soft-ware. Viewing was binocular, but only the right eye was mon-itored. A chin rest and forehead rest were used to minimizehead movements. Following calibration, gaze-position error wasless than 0.5◦. The chess problems were presented using images(755 × 755 pixels) that were created using standard chess soft-ware (Chessbase 11). These images were displayed on a 21 in.ViewSonic monitor with a refresh rate of 150 Hz and a screenresolution of 1024 × 768 pixels. Participants were seated 60 cmfrom the monitor, and the width of one square on the chessboardequaled approximately 3.4 degrees of visual angle.

Prior to the experiment, the participants were instructed tochoose white’s best move as quickly and as accurately as possi-ble, and they were told that they would be given a maximumof 3 min to respond to each problem. At the start of each trial,the participants were required to look at a fixation point in thecenter of the screen, prior to the presentation of the chessboard.The participants were asked to press a button as soon as they hadmade their decision, and they then reported their move verballyto the experimenter. If 3 min elapsed prior to the button press(this occurred on less than 1% of the experimental trials for thenovices, and 0% of trials for the experts), then the chessboard wasremoved from the screen and the chess player was prompted toimmediately provide their best answer.

RESULTSOur main goal was to use eye movements to examine expertvs. novice differences in how attention was allocated to the

relevant and irrelevant regions of the chessboard. However, priorto reporting the eye movement results, we will first report severalglobal measures of performance (i.e., accuracy, reaction times) asa function of the chess player’s level expertise (expert, novice).

To assess move quality, we asked an international chess mas-ter who did not participate in the study to rate the quality ofeach move on a scale from 1 to 10 (1 = a blunder, 10 = one ofthe best moves on the board), and we consulted the evaluationfunction from a chess engine (Houdini 2 Pro), which provides ascore (expressed in pawn units) to quantify the change in White’spositional advantage as a result of the move chosen. For both ofthese measures of accuracy, the experts showed superior perfor-mance than the novices (df = 39; all ts > 2.0, all ps < 0.05).Specifically, the average move quality rating was 9.5 (SE = 0.12)for the experts and 6.9 (SE = 0.26) for the novices, and the aver-age chess engine score was 2.1 (SE = 0.17) for the experts and1.5 (SE = 0.16) for the novices. Moreover, the experts selectedthe best move on the board (i.e., a move that received a rat-ing of 10), on an average of 93 % of trials (SE = 2%), whereasthe novices selected the best move on an average of 52% oftrials (SE = 5%), and this expert/novice difference in accuracywas significant: t(39) = 7.12, p < 0.001. In addition, there wasa marginally significant trend toward faster reaction times forthe experts (M = 28, 946 ms; SE = 5441 ms) than for the novices(M = 41, 854 ms; SE = 4748 ms), t(39) = 1.78, p = 0.084. Moreinterestingly, within the expert group there was a negative correla-tion between the mean reaction time of each player and their Elorating (r = −0.617, p < 0.01), which indicates that increases inchess rating were associated with faster performance. Finally, theexperts’ (but not the novices’) reaction times were significantlyfaster when the relevant region was on the right rather than theleft side of the board, t(16) = 2.16, p < 0.05 (for similar findings,see De Groot and Gobet, 1996).

Next, we analyzed eye movements to examine the extent towhich the expert and novice chess players directed their atten-tion toward the relevant and irrelevant regions of the board.For all of the eye movement analyses reported below, we ana-lyzed correct trials only (i.e., trials that elicited a 10-rated move),to ensure that the experts and novices were matched for accu-racy. Given our interest in the time-course of relevancy effects,we began our analysis by examining an early measure of pro-cessing (i.e., first-dwell duration, which is the duration of thefirst dwell on a given region, where a dwell is defined as oneor more consecutive fixations on the target region, prior to theeyes moving to a different region of the board) as well as a latermeasure of processing (i.e., total time, which is the sum of theduration of all of the dwells on a region for the entire trial).To explore the pattern of results for each measure, we exam-ined 2 × 2 ANOVAs that included relevancy (relevant, irrelevant)as a within-subjects factor, and expertise (expert, novice) as abetween-subjects factor. For both the first-dwell and total timemeasures, there was a main effect of relevancy (i.e., longer dwellson relevant than irrelevant regions), all Fs > 8, all ps < 0.01,and a main effect of expertise (i.e., longer dwells for novices thanexperts), all Fs > 60, all ps < 0.001. More importantly, there was asignificant interaction for the first-dwell measure [F(1, 39) = 4.38,p < 0.05], but not for the total time measure (F < 1). As can

www.frontiersin.org August 2014 | Volume 5 | Article 941 | 13




be seen from Figures 1A,B, this interesting dissociation betweenthese two measures occurred because the first dwell measure pro-duced significant relevancy effects for the experts [t(16) = 3.51,p < 0.01] but not for the novices (t < 1), whereas the total timemeasure produced relevancy effects for both groups (all ts > 2,all ps < 0.05). Interestingly, the interaction between expertise andrelevancy for the first-dwell measure was solely due to an increasein the number of fixations for relevant compared to irrelevantregions for the experts [relevant: M = 2.19, SE = 0.21, irrele-vant: M = 1.39, SE = 0.07, t(16) = 3.95, p < 0.01] but not forthe novices (relevant: M = 2.05, SE = 0.16, irrelevant: M = 1.98,SE = 0.23, t < 1) as shown by a significant interaction, F(1, 39) =8.28, p < 0.01. In contrast, the mean fixation duration for thefirst dwell did not vary as a function of relevancy or expertise (allFs < 1).

The above first-dwell findings imply that experts are fasterthan the novices at detecting relevant information. To further

FIGURE 1 | The duration of the First Dwell (A) Total Time (B), and the

Cumulative Time of the first five ordinal fixations in the trial (C) as a

function of relevancy (relevant vs. irrelevant) and level of expertise

(expert, novice).

explore this effect, we analyzed the first five fixations of thetrial to quantify the amount of time that experts and novicesspent fixating the relevant and irrelevant regions at the start ofthe trial (a similar analysis of the first five fixations was con-ducted by Charness et al., 2001). Specifically, for each fixationposition ranging from one to five (fixation position one corre-sponded to the fixation which began following the initial saccadein the trial), we calculated the cumulative sum of all of the fixa-tions on the relevant and irrelevant regions up to and includingthe current fixation position. This analysis was conducted sepa-rately for each participant and each condition (i.e., relevant vs.irrelevant), and then averaged across participants to produce thefigure shown in Figure 1C. As can be seen from this figure, theexperts showed stronger and earlier relevancy effects than thenovices. This pattern of results was reflected in a three-way inter-action, F(4,36) = 2.74, p < 0.05 when we examined a 5 × 2 × 2ANOVA that included fixation position (1,2,3,4,5) and relevancy(relevant, irrelevant) as within-subjects variables, and expertise(expert, novice) as a between-subjects variable. Consistent withthis three-way interaction, the experts showed a significant rele-vancy effect [F(1, 16) = 6.64, p < 0.05] that became stronger overtime as shown by a relevancy by time interaction [F(4, 13) = 6.67,p < 0.01], whereas the novices did not show a relevancy effect oran interaction (all Fs < 1).

Taken together, the first-dwell findings and the cumulativetime analyses indicate that experts are faster at detecting relevantinformation than novices, which supports the notion that chessexpertise reflects an advantage in encoding chess-related visualconfigurations (De Groot, 1946, 1965; Chase and Simon, 1973a,b;De Groot and Gobet, 1996; Charness et al., 2001; Reingoldet al., 2001a,b; Bilalic et al., 2010a, 2011a,b, 2012; for reviewssee Reingold and Charness, 2005; Reingold and Sheridan, 2011).During the present study, this perceptual processing advantageenabled skilled players to rapidly focus on chess configurationsthat were relevant to the best move on the board.

DISCUSSIONThe present experiment examined the time-course and magni-tude of relevancy effects on expert and novice chess players’ eyemovements during a choose-a-move task. Our most importantfinding was that the eye movements of the experts, but not thenovices, revealed a rapid differentiation between regions of thechess board that were relevant vs. irrelevant to the best moveon the board. Specifically, the experts, but not the novices, spentmore time looking at relevant than irrelevant regions during theearly part of the trial (i.e., during the first-dwell on a region,and during the first five fixations of the trial), whereas both theexperts and novices showed strong relevancy effects later on in thetrial. Importantly, these findings were obtained using an experi-mental paradigm that provided a well-controlled manipulationof relevancy, such that the relevant and irrelevant regions of theboard were counterbalanced across participants (see Appendix Ain Supplementary material).

Similar to the present findings, prior studies have also shownenhanced relevancy detection by chess experts (Tikhomirov andPoznyanskaya, 1966; Simon and Barenfeld, 1969; De Groot andGobet, 1996; Charness et al., 2001; Bilalic et al., 2010a, 2012).






Moreover, this relevancy detection advantage is a particularinstance of the perceptual encoding advantage that has beenshown by chess experts in a variety of tasks employing domain-related stimuli (De Groot, 1946, 1965; Chase and Simon, 1973a,b;De Groot and Gobet, 1996; Charness et al., 2001; Reingoldet al., 2001a,b; Bilalic et al., 2010a, 2011a,b, 2012; for reviewssee Reingold and Charness, 2005; Reingold and Sheridan, 2011).To explain this perceptual encoding advantage, Chase and Simon(1973a,b) proposed that chess expertise develops due to exten-sive practice with domain-related visual-configurations. Over thecourse of thousands of hours of practice, chess experts store mem-ories for configurations of chess pieces in memory in the formof chunks (Chase and Simon, 1973a,b), which are supplementedby larger memory structures called templates (Gobet and Simon,1996, 2000). These memory structures lead to a perceptual encod-ing advantage such that chess experts are able to process chessstimuli in terms of larger configurations of pieces, rather thanindividual features. Consequently, chess players are able to usetheir memory for chess configurations to guide their search forthe best move on the board, rather than exhaustively searchingall possible moves. This theoretical account is consistent with thepresent study’s finding that the chess experts were able to rapidlyfocus on information that was relevant to the best move on theboard.

Beyond the chess domain, the present findings are also con-sistent with findings from other domains concerning the impor-tance of perceptual processing during skilled performance. Asreviewed by Reingold and Sheridan (2011), experts in manydomains of expertise have been shown to efficiently processdomain-related material in terms of larger configurations, asshown by findings that radiologists can rapidly fixate abnormal-ities in less than a second (Kundel et al., 2008). Moreover, thiskey role of perceptual processing in expertise coincides with otherevidence for perceptual specificity effects in memory and learn-ing (for reviews, see Levy, 1993; Roediger and McDermott, 1993;Reingold, 2002), such as recent findings that eye fixation timesare shorter for words that were read twice in the same typography(i.e., font) rather than in two different typographies (Sheridanand Reingold, 2012a,b), and findings that chess experts performbetter when viewing familiar chess symbols compared to a condi-tion in which letters (i.e., B = Bishop, P = Pawn, etc.) were showninstead of the symbols (Reingold et al., 2001a).

More generally, the relevancy effects from the present studyadd to the growing evidence that higher level cognitive processescan rapidly influence eye movement control (e.g., the durationand location of fixations) on a variety of tasks. In fact, ever sincethe seminal eye tracking work by Yarbus (1967), it has been well-known that our eye movements are biased toward aspects of avisual image that are relevant to our current goals, and awayfrom areas that are irrelevant. As further evidence for the roleof higher cognitive processing in guiding eye movements, par-ticipants in visual search studies spend more time fixating ondistractors that are related (e.g., visually similar) rather than unre-lated to the target (e.g., Findlay and Gilchrist, 1998; Reingold andGlaholt, 2014), participants in face perception tasks preferentiallylook at relevant features, such as the eyes (e.g., Henderson et al.,2005), participants in scene perception tasks spend more time

fixating information that is task-relevant rather than irrelevant(Glaholt and Reingold, 2012), and the eye movements of skilledreaders reveal rapid effects of a variety of higher-level lexical, lin-guistic and cognitive variables (e.g., Rayner, 1998, 2009; Staubet al., 2010; Staub, 2011; Reingold et al., 2012; Sheridan andReingold, 2012c,d). Taken together, these findings lend supportto models of eye movement control that predict a strong eye-mind link, such that ongoing cognitive processing can have a rapideffect on lower-level perceptual and oculomotor processing (forrecent reviews, see Reingold et al., 2012, in press). Moreover, thesefindings underscore that skilled performance reflects a complexinter-play of perceptual and cognitive processing, and future workcan examine the extent to which similar findings from multipledomains are reflective of common underlying mechanisms.

Finally, a key contribution of the present study is that weintroduced a new experimental paradigm to provide a carefullycontrolled manipulation of relevancy. As shown in Appendix Ain Supplementary material, we created two counterbalanced ver-sions of each chess problem that differed by a single piece, andthe regions that were relevant in one version were irrelevant inthe other version (and vice versa). Given that the relevant andirrelevant regions of the board were well-matched, we can con-clude that the relevancy effects in the experiment were solely dueto the relevancy of a given region to the best move, and not tosome other confound (e.g., differences in visual saliency, locationon the board, number of pieces in the region, etc.). The presentparadigm could be used in the future to investigate additionaltopics concerning the impact of relevancy on a variety of aspectsof chess expertise.

ACKNOWLEDGMENTSWe are especially grateful to all of the chess players who partic-ipated in the experiment, and to Rick Lahaye for his valuableassistance with stimuli creation. This research was supported byan NSERC grant to Eyal Reingold, a Postdoctoral Fellowship(PDF) awarded to Heather Sheridan from NSERC, and supportfrom the Centre for Vision and Cognition (CVC), University ofSouthampton, UK.

SUPPLEMENTARY MATERIALThe Supplementary Material for this article can be found onlineat: http://www.frontiersin.org/journal/10.3389/fpsyg.2014.

00941/abstract

REFERENCESBilalic, M., Kiesel, A., Pohl, C., Erb, M., and Grodd, W. (2011a). It takes two–

skilled recognition of objects engages lateral areas in both hemispheres. PLoSONE 6:e16202. doi: 10.1371/journal.pone.0016202

Bilalic, M., Langner, R., Erb, M., and Grodd, W. (2010a). Mechanisms and neuralbasis of object and pattern recognition: a sudy with chess experts. J. Exp. Psychol.Gen. 139, 728–742. doi: 10.1037/a0020756

Bilalic, M., Langner, R., Ulrich, R., and Grodd, W. (2011b). Many faces of expertise:fusiform face area in chess experts and novices. J. Neurosci. 31, 10206–10214.doi: 10.1523/JNEUROSCI.5727-10.2011

Bilalic, M., McLeod, P., and Gobet, F. (2010b). The mechanism of the Einstellung(set) effect: a pervasive source of cognitive bias. Curr. Dir. Psychol. Sci. 19,111–115. doi: 10.1177/0963721410363571

Bilalic, M., McLeod, P., and Gobet, F. (2008a). Inflexibility of experts - reality ormyth? Quantifying the Einstellung effect in chess masters. Cogn. Psychol. 56,73–102. doi: 10.1016/j.cogpsych.2007.02.001







Bilalic, M., McLeod, P., and Gobet, F. (2008b). Why good thoughts block betterones: the mechanism of the pernicious Einstellung (set) effect. Cognition 108,652–661. doi: 10.1016/j.cognition.2008.05.005

Bilalic, M., Turella, L., Campitelli, G., Erb, M., and Grodd, W. (2012). Expertisemodulates the neural basis of context dependent recognition of objects and theirrelations. Hum. Brain Mapp. 33, 2728–2740. doi: 10.1002/hbm.21396

Charness, N. (1992). The impact of chess research on cognitive science. Psychol.Res. 54, 4–9. doi: 10.1007/BF01359217

Charness, N., Reingold, E. M., Pomplun, M., and Stampe, D. M. (2001). The per-ceptual aspect of skilled performance in chess: evidence from eye movements.Mem. Cogn. 29, 1146–1152. doi: 10.3758/BF03206384

Chase, W. G., and Simon, H. A. (1973a). Perception in chess. Cogn. Psychol. 4,55–81. doi: 10.1016/0010-0285(73)90004-2

Chase, W. G., and Simon, H. A. (1973b). “The mind’s eye in chess,” in VisualInformation Processing, ed W. G. Chase (New York, NY: Academic Press),215–281.

De Groot, A. D. (1946). Het Denken Van Den Schaker. Amsterdam: NoordHollandsche.

De Groot, A. D. (1965). Thought and Choice in Chess. The Hague: Mouton.De Groot, A. D., and Gobet, F. (1996). Perception and Memory in Chess. Assen: Van

Gorcum.Elo, A. E. (1965). Age changes in master chess performances. J. Gerontol. 20,

289–299.Elo, A. E. (1986). The Rating of Chessplayers, Past and Present, 2nd Edn. New York,

NY: Arco chess.Findlay, J. M., and Gilchrist, I. D. (1998). “Eye guidance and visual search” in Eye

Guidance in Reading, Driving and Scene Perception, ed G. Underwood (Oxford:Elservier), 295–312.

Glaholt, M. G., and Reingold, E. M. (2012). Direct control of fixation times in sceneviewing: evidence from analysis of the distribution of first fixation duration. Vis.Cogn. 20, 605–626. doi: 10.1080/13506285.2012.666295

Gobet, F., and Charness, N. (2006). “Expertise in chess” in The CambridgeHandbook of Expertise and Expert Performance, eds K. A. Ericsson, N. Charness,P. J. Feltovich, and R. R. Hoffman (New York, NY: Cambridge University Press),523–538.

Gobet, F., and Simon, H. A. (1996). Templates in chess memory: a mechanism forrecalling several boards. Cogn. Psychol. 31, 1–40. doi: 10.1006/cogp.1996.0011

Gobet, F., and Simon, H. A. (2000). Five seconds or sixty? Presentation time inexpert memory. Cogn. Sci. 24, 651–682. doi: 10.1207/s15516709cog2404_4

Henderson, J. M., Williams, C. C., and Falk, R. J. (2005). Eye movements arefunctional during face learning. Mem. Cogn. 33, 98–106. doi: 10.3758/BF03195300

Kundel, H. L., and Nodine, C. F. (1975). Interpreting chest radiographs withoutvisual search. Radiology 116, 527–532. doi: 10.1148/116.3.527

Kundel, H. L., Nodine, C. F., Krupinski, E. A., and Mello-Thoms, C. (2008). Usinggaze-tracking data and mixture distribution analysis to support a holistic modelfor the detection of cancers on mammograms. Acad. Radiol. 15, 881–886. doi:10.1016/j.acra.2008.01.023

Levy, B. A. (1993). “Fluent rereading: an implicit indicator of reading skill devel-opment,” in Implicit Memory: New directions in Cognition, Development, andNeuropsychology, eds P. Graf and M. Masson (Hillsdale, NJ: Lawrence ErlbaumAssociates Inc.), 49–73.

Rayner, K. (1998). Eye movements in reading and information processing: 20 yearsof research. Psychol. Bull. 124, 372–422. doi: 10.1037/0033-2909.124.3.372

Rayner, K. (2009). Eye movements in reading: models and data. J. Eye Mov. Res. 2,1–10. doi: 10.1017/S0140525X03520106

Reingold, E. M. (2002). On the perceptual specificity of memory representations.Memory 10, 365–379. doi: 10.1080/09658210244000199

Reingold, E. M., and Charness, N. (2005). “Perception in chess: evidence from eyemovements,” in Cognitive Processes in Eye Guidance, ed G. Underwood (Oxford:Oxford University Press), 325–354.

Reingold, E. M., Charness, N., Pomplun, M., and Stampe, D. M. (2001a). Visualspan in expert chess players: evidence from eye movements. Psychol. Sci. 12, 48.doi: 10.1111/1467-9280.00309

Reingold, E. M., Charness, N., Schultetus, R. S., and Stampe, D. M.(2001b). Perceptual automaticity in expert chess players: parallel encoding

of chess relations. Psychon. Bull. Rev. 8, 504–510. doi: 10.3758/BF03196185

Reingold, E. M., and Glaholt, M. G. (2014). Cognitive control of fixation durationin visual search: the role of extrafoveal processing. Vis. Cogn. 22, 610–634. doi:10.1080/13506285.2014.881443

Reingold, E. M., Reichle, E. D., Glaholt, M. G., and Sheridan, H. (2012). Directlexical control of eye movements in reading: evidence from a survival analysisof fixation durations. Cogn. Psychol. 65, 177–206. doi: 10.1016/j.cogpsych.2012.03.001

Reingold, E. M., and Sheridan, H. (2011). “Eye movements and visual exper-tise in chess and medicine,” in Oxford Handbook on Eye Movements, eds S. P.Liversedge, I. D. Gilchrist, and S. Everling (Oxford: Oxford University Press),528–550.

Reingold, E. M., Sheridan, H., and Reichle, E. D. (in press). “Direct lexical and non-lexical control of fixation duration in reading,” in Oxford Handbook on Reading,eds A. Pollatsek and A. Treisman (Oxford: Oxford University Press).

Roediger, H. L., and McDermott, K. B. (1993). “Implicit memory in normal humansubjects,” in Handbook of Neuropsychology, Vol. 8. eds H. Spinnler and F. Boller(Amsterdam: Elsevier), 63–131.

Sheridan, H., and Reingold, E. M. (2012a). Perceptual specificity effects inrereading: evidence from eye movements. J. Mem. Lang. 67, 255–269. doi:10.1016/j.jml.2012.05.005

Sheridan, H., and Reingold, E. M. (2012b). Perceptually specific and percep-tually non-specific influences on rereading benefits for spatially transformedtext: evidence from eye movements. Conscious. Cogn. 21, 1739–1747. doi:10.1016/j.concog.2012.10.002

Sheridan, H., and Reingold, E. M. (2012c). The time course of contextual influ-ences during lexical ambiguity resolution: evidence from distributional analysesof fixation durations. Mem. Cogn. 40, 1122–1131. doi: 10.3758/s13421-012-0216-2

Sheridan, H., and Reingold, E. M. (2012d). The time course of predictability effectsin reading: evidence from a survival analysis of fixation durations. Vis. Cogn. 20,733–745. doi: 10.1080/13506285.2012.693548

Sheridan, H., and Reingold, E. M. (2013). The mechanisms and boundary con-ditions of the Einstellung effect in chess: evidence from eye movements. PLoSONE 8:e75796. doi: 10.1371/journal.pone.0075796

Simon, H. A., and Barenfeld, M. (1969). Information-processing analysis ofperceptual processes in problem solving. Psychol. Rev. 76, 473–483. doi:10.1037/h0028154

Staub, A. (2011). Word recognition and syntactic attachment in reading: evi-dence for a staged architecture. J. Exp. Psychol. Gen. 140, 407–433. doi:10.1037/a0023517

Staub, A., White, S. J., Drieghe, D., Hollway, E. C., and Rayner, K. (2010).Distributional effects of word frequency on eye fixation durations. J. Exp.Psychol. Hum. Percept. Perform. 36, 1280–1293. doi: 10.1037/a0016896

Tikhomirov, O. K., and Poznyanskaya, E. (1966). An investigation of visual searchas a means of analyzing heuristics. Sov. Psychol. 5, 2–15.

Yarbus, A. L. (1967). Eye Movements and Vision. New York, NY: Plenum Press.

Conflict of Interest Statement: The authors declare that the research was con-ducted in the absence of any commercial or financial relationships that could beconstrued as a potential conflict of interest.

Received: 02 April 2014; paper pending published: 08 June 2014; accepted: 06 August2014; published online: 25 August 2014.Citation: Sheridan H and Reingold EM (2014) Expert vs. novice differences in thedetection of relevant information during a chess game: evidence from eye movements.Front. Psychol. 5:941. doi: 10.3389/fpsyg.2014.00941This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Sheridan and Reingold. This is an open-access article distributedunder the terms of the Creative Commons Attribution License (CC BY). The use, dis-tribution or reproduction in other forums is permitted, provided the original author(s)or licensor are credited and that the original publication in this journal is cited, inaccordance with accepted academic practice. No use, distribution or reproduction ispermitted which does not comply with these terms.














doi: 10.3389/fpsyg.2014.00949

Expert and competent non-expert visual cues duringsimulated diagnosis in intensive careClare McCormack 1, Mark W. Wiggins1*, Thomas Loveday1 and Marino Festa 2

1 Centre for Elite Performance, Expertise, and Training, Macquarie University, North Ryde, NSW, Australia2 Paediatric Intensive Care Unit, Kim Oates Australian Paediatric Simulation Centre, Children’s Hospital at Westmead, Westmead, NSW, Australia

Edited by:

Guillermo Campitelli, Edith CowanUniversity, Australia

Reviewed by:

Guillermo Campitelli, Edith CowanUniversity, AustraliaMerim Bilalic, Alpen-Adria UniversitätKlagenfurt, Austria

*Correspondence:

Mark W. Wiggins, Centre for ElitePerformance, Expertise, and Training,Macquarie University, Balaclava Road,North Ryde, NSW 2109, Australiae-mail: [email protected]

The aim of this study was to examine the information acquisition strategies of expertand competent non-expert intensive care physicians during two simulated diagnosticscenarios involving respiratory distress in an infant. Specifically, the information acquisitionperformance of six experts and 12 competent non-experts was examined using an eye-tracker during the initial 90 s of the assessment of the patient. The results indicatedthat, in comparison to competent non-experts, experts recorded longer mean fixations,irrespective of the scenario. When the dwell times were examined against specific areasof interest, the results revealed that competent non-experts recorded greater overall dwelltimes on the nurse, where experts recorded relatively greater dwell times on the head andface of the manikin. In the context of the scenarios, experts recorded differential dwelltimes, spending relatively more time on the head and face during the seizure scenario thanduring the coughing scenario. The differences evident between experts and competentnon-experts were interpreted as evidence of the relative availability of task-specific cues orheuristics in memory that might direct the process of information acquisition amongstexpert physicians. The implications are discussed for the training and assessment ofdiagnostic skills.

Keywords: expertise, cue utilization, diagnosis, medicine, simulation

INTRODUCTIONThe accurate initial assessment of clinical patients in time-criticalemergencies is an essential component of timely and appropriateintervention by critical care teams (Pham et al., 2012). It mit-igates the further deterioration of the patient’s condition andpotentially reduces mortality and the additional burden on analready strained healthcare system. Nevertheless, it is a processthat occurs within a short time-period and with potentially min-imal information, thereby increasing the likelihood of error (Elyet al., 2011).

On the basis that assessments are required within a relativelyshort period and with minimal information, it is likely that aphysician will engage lean and rapid cognitive strategies suchas satisficing, relying on productions or relationships betweenpatterns of information to guide the initial process of diagno-sis (Simon, 1972; Marewski and Gigerenzer, 2012). Productionscomprise rules-of-thumb or condition-action (IF-THEN) state-ments that are resident in memory and that can be used to assistthe interpretation of a situation or event (Anderson, 1982; Hamm,2014). For example, in the medical context, IF a patient presentswith an elevated temperature, THEN it is normally associated withthe presence of an infection.

The development and application of productions is gener-ally associated with a reduction in cognitive load, since theirapplication obviates the requirement for compensatory strategiesthat require the retention of task-related information in workingmemory (Sweller, 1988). However, such rules-of-thumb are, bydefinition, not necessarily applicable in all situations, and there are

many cases where the application of productions has been asso-ciated with the commission of errors (Croskerry, 2003; Normanet al., 2014).

The acquisition of information as a prelude to the diagnosis ofa particular condition is based, in part, upon the features that areimmediately apparent on presentation to the physician (Croskerry,2009a; Stolper et al., 2011). Where an association exists in memory,a feature or combination of features is presumed to trigger a pro-duction that will be interpreted as the basis of a diagnosis or willprovide the impetus for the acquisition of additional informationnecessary to form a diagnosis (Khader et al., 2011). This processis consistent theoretically with the initial stages of recognition-driven decision-making where the condition-action statementsthat comprise productions are referred to as cues (Klein, 2008).

The acquisition and application of cues is thought to explainthe rapid and consistently accurate behavior of genuine experts(Mann et al., 2007; Kahneman and Klein, 2009). In the context ofthe Recognition-Primed Decision model, cues trigger associationsin memory that subsequently provide the basis for mental sim-ulations that, in turn, guide a response (Klein, 2008). Brunswik(1955), in his Lens Model, also proposes that the likelihood ofan association being triggered is dependent upon the frequencywith which features in the environment match features in mem-ory. Finally, Stokes et al. (1997) incorporate cues as the precursorto diagnosis in their theoretical model of expert decision-makingin the aviation context.

Like productions, cues are essentially feature-event/objectrelationships in memory that enable the rapid assessment of


http://www.frontiersin.org/Psychology/





http://www.frontiersin.org/Journal/10.3389/fpsyg.2014.00949/abstract








McCormack et al. Visual cues in diagnosis

a situation and, subsequently, the formulation of a response(Wiggins, 2006, 2012). Establishing the existence of cues has gen-erally been inferred on the basis of responses to domain-specificstimuli. For example, Morrison et al. (2013) demonstrated that,in comparison to non-experts, expert forensic investigators wererelatively consistent and responded more rapidly in assessing therelatedness of feature/event pairs relating to a murder investi-gation. Similarly, Wiggins and O’Hare (1995) established thatthe acquisition of weather-related information differed betweenexperts and non-expert pilots, with the former being less likely toaccess information in the sequence in which it was presented. Thisbehavior has been interpreted as evidence to suggest a greater levelof cue utilization amongst experts.

The association between levels of cue utilization and expertisehas been established in squash (Abernethy, 1990), power con-trol (Loveday et al., 2013a), pediatric assessment (Loveday et al.,2013b), and aviation (Wiggins et al., 2014). Measures of cue uti-lization have also differentiated performance in the context ofsoftware engineering (Loveday and Wiggins,2014). However, theseapproaches have been based on generalized behavior and there isno indication as to the specific cues involved and how they mightbe activated in response to the presence of features.

As experts gain experience within a particular context,Anderson (1982) suggests that productions are revised so that theybecome more precise and discriminate between different circum-stances. Referred to as discrimination, it is a process that coincideswith generalization where it becomes evident that a particularproduction is equally applicable across a range of conditions. Thiscombination of discrimination and generalization may explainboth the domain specificity of experts, together with their capacityto perceive underlying similarities between situations (Shanteau,1988).

If experts possess a highly refined repertoire of task-relatedcues in memory, then the immediate features associated with twodiagnostic scenarios that differ in their immediate features butincorporate a similar intrinsic etiology, should trigger the bottom-up application of distinct cues, and these differences should beevident in differences in the process of information acquisition(Patel and Groen, 1986; Croskerry, 2009b). Empirical support forthis capacity for bottom-up discrimination can be drawn fromresearch into the Einstellung Effect in which visual attention dur-ing expert problem-solving is implicitly drawn toward familiarsolutions, even at the expense of novelty (Bilalic et al., 2010). Sincecompetent non-experts have yet to develop highly specialized cues,they are not expected to alter their information acquisition inresponse to the differences in the immediate features of the task.

MATERIALS AND METHODSPARTICIPANTSThe participants in the present study were drawn from a conve-nience sample of medical practitioners of different levels of gradeand seniority, working in the pediatric and neonatal intensive careunits of a tertiary children’s hospital. The local research and ethicscommittee approved the study, and individual participant consentwas obtained from each clinician examined.

The participants comprised 11 male and seven female physi-cians employed in either pediatric or neonatal intensive care.

Their mean age was 40.5 years, SD = 10.6. Establishing thatexpertise, rather than experience, has been acquired, requiresthat some formal criterion be established that is typically basedon measures of performance. In the context of medical prac-titioners, an indirect measure of expertise is the seniority oftheir role (Patel and Groen, 1991). This constitutes recogni-tion that they have successfully attained a level of performancewhere their medical interventions are both accurate and consis-tent over an extended period of time. While a positive relationshipwill inevitably exist between years of accumulated experienceand performance, the recognition of expertise amongst peerspresumes that a level of performance has been reached that isexceptional in comparison to other practitioners (Loveday andWiggins, 2014). Therefore, consistent with this perspective, theparticipants were classified as expert or competent non-expertsbased on their occupational position (consultant/staff special-ist, n = 6, or trainee registrar/fellow, n = 12) according tothe criteria established by Patel and Groen (1991). Their accu-mulated experience working in medicine was between 6 and42 years, m = 16.5, SD = 10.6, with a range of 1–35 years inthe intensive care environment, m = 9.9, SD = 10.3. Expertsrecorded a mean 23.0 years experience working in intensivecare, SD = 9.33, compared to a mean 4.8 years for competentnon-experts, SD = 4.57.

SIMULATIONA realistic scenario and naturalistic environment was createdfor the study by using a high-fidelity infant manikin (LaerdalSimBaby) connected to a monitor that displayed simulated phys-iological parameters and appropriate corresponding alarms andsounds in situ in an intensive care cot in a bedspace within thepediatric intensive care unit of a tertiary children’s hospital. Theconfiguration of the room was typical of a bedspace in the pediatricintensive care unit and was familiar to all study participants (seeFigure 1). A nasogastric feeding tube was inserted and attached to acontinuous feeding pump with enteral feed attached. The manikinwas also connected to nasal prong oxygen with a wall-mountedoxygen flow-meter, an intravenous drip via a peripherally insertedintravenous cannula, and an appropriately sized blood pressurecuff was attached to the right arm. A familiar and experiencedpediatric intensive care nurse with a pre-scripted dialog was usedas a confederate actor within the scenario.

An IVIEW XTM HED eye tracking system manufactured bySensoMotoric Instruments was used to record the eye movementsof participants, in addition to scene video and audio recording.The system consists of a fully mobile, head-mounted device withtwo cameras attached, one recording the scene and one trainedon the participant’s eye, recording gaze and pupil data. A pieceof clear plastic was fixed in front of one eye. The device was con-nected to a notebook computer, which powered the cameras andstored gaze, video and audio data. The gaze sampling rate usedwas 50 Hz, and a fixation was defined as 100 ms with maximumdispersion of 20 pixels. Based on the limitations imposed by framerate acquisition and the need to include all features in the intensivecare environments, features were broadly categorized as belong-ing to one of six areas of interest (AOI). Each area of interest wasdefined by anatomical or environmental relationships.


http://www.frontiersin.org/Cognition/



FIGURE 1 | Schematic showing the location of the various areas of

interest (AoI), relative to the participant.

SCENARIOSThe two scenarios used during the study were written by twosubject-matter experts, both of whom were senior intensive carespecialists working in the pediatric intensive care unit. The sce-narios were designed around two immediate features, the first ofwhich related to the head and face of the manikin. In particu-lar, the level of consciousness of the child would be an importantdeterminant in the seizure scenario, but would be less significant inthe context of the coughing scenario. This information would bedetermined through the child’s facial features, including the eyes.In the coughing scenario, the information provided spontaneouslyby the assisting nurse was the immediate feature, since this wouldbe an important determinant as to whether any respiratory assis-tance had been provided. Participants were randomly allocated toeither the coughing or the seizure scenario as their first scenario,and all participants completed both scenarios.

The initial disease state was identical for both scenarios withthe immediate features becoming evident as the symptomatologyemerged. A simple respiratory arrest scenario in a self-ventilatingmonitored patient was used as the initial disease state, since itavoided potentially confounding effects that might be introducedby complex or unfamiliar equipment.

In the first minute of the coughing scenario, the patient demon-strated a heart rate of 150 beats per minute (BPM), blood pressureof 77/40 and a respiratory rate of 66 breaths per minute. Satura-tion was at 94% on 1 liter/min of nasal prong oxygen with goodconnections. This information, and electrocardiography (ECG),was displayed on the monitor. The patient showed see-saw breath-ing with bilateral crackles as well as grunting that was cycling withbreaths. The cot was tilted at 30◦. The scenario began with thenurse introducing the patient, saying: “The ward is about ready totake this baby with bronchiolitis, but I’m concerned about whetherhe’s OK to be discharged from PICU as he’s had a couple of shortdesaturations as I’ve been looking after him this morning.” They

were also advised that the patient presented to the emergencydepartment the previous evening with increased respiratory work,and was found to have respiratory syncytial virus (RSV) – pos-itive bronchiolitis, and hyperinflation shown on a chest x-ray.The patient was admitted to PICU late on the previous after-noon for possible continuous positive airway pressure therapy,but improved with nasal oxygen. Feeds were started at 6.00amthat morning, but had been stopped a few hours later following asecond desaturation episode. Desaturations were associated withcoughing and not with vomiting or the reflux of feed. No apnoea,bradycardia, or seizure was noted at the time. The temperaturewas at 37.6◦C, and the patient was not on antibiotics. A pertussisswab had not been taken. A full blood count on admission showedhemoglobin (Hb) of 10.7 g per deciliter, white cell count (WCC)was 9.3 cells per cubic millimeter (Neutrophils 5.3, Leukocytes4.0), and platelets at 210 cells per cubic milliliter.

After 1 min had elapsed, the manikin was set to cough for 20 s,desaturate to 84% over 40 s, and become bradycardic to 104 bpmover 40 s. At this point, the nurse prompted participants, say-ing: “This is what he did before you came in.” After 2 min and10 s, saturation increased to 99% if the participant had used anoxygen bag, or to 94% if no adjustment to oxygen administra-tion was made. Heart rate increased to 160 over 20 s, and thepatient showed grunting and see-saw rasps as had occurred previ-ously. The scenario concluded following a duration of 3 min and30 s.

Prior to commencing the second scenario, participants wereadvised that this was a “new patient,” not related to the previousscenario. In the first minute of the seizure scenario, the patient hada heart rate of 120 bpm, blood pressure of 99/70 and respiratoryrate of 33 breaths per minute. Saturation was at 94% on 1 liter/minof nasal prong oxygen with good connections. This information,and ECG, was displayed on the monitor. The patient showed see-saw breathing with bilateral crackles as well as grunting that wascycling with breaths. The cot was tilted at 30◦ and the scenariobegan with the nurse introducing the patient, saying: “This babyhas just been brought up from the ward by the nurse practitioneras he has had a couple of episodes of desaturation with stiffening ofhis arms and legs on the ward. I’m a little bit worried about himas he’s just had another similar episode and dropped his ‘sats’ to themid 80 s. I’ve just done a capillary gas, which is in the gas machinenow.”

The participants were also advised that the patient was a 6 weekold baby delivered at full term with no neonatal problems. Thepatient was presented to the ward 2 days previously with RSV pos-itive bronchiolitis and hyperinflation shown on a chest x-ray. Sincethen, the patient had been on full maintenance intravenous fluids(N/4 and 5% dextrose) and nil by mouth. The patient was admittedto PICU an hour earlier. Since then, he had shown desaturationto the mid 85 associated with unusual movements of the torsoand stiffening of limbs, and an increase in heart rate. The desat-urations would self-correct after a minute of nasal prong oxygen,increased to 2 liter per minute. The temperature was at 37.6◦C,and the patient was not on antibiotics. A pertussis swab had notbeen taken. A full blood count on admission showed Hb of 10.7 gper deciliter, WCC was 9.3 cells per cubic millimeter (Neutrophils5.3, Leukocytes 4.0), and platelets at 210 cells per cubic milliliter.





After 1 min had elapsed, the manikin was programmed to showrapid and slow torso movements over 20 s, desaturation to 84%over 40 s and tachycardia to 180 over 40 s. At this point, the nurseprompted participants, saying: “This is what he did before you camein. Here’s the cap gas (hands over blood gas analysis).” After 2 minand 10 s, saturation increased to 99% over 20 s if the participanthad used an oxygen bag, or to 94% if no adjustment to oxygenadministration was made. Heart rate dropped to 160 over 20 s,and the patient showed see-saw rasps with the respiratory ratestill at 33 breaths per minute. The scenario concluded following aduration of 3 min and 30 s.

PROCEDUREThe participants completed a pre-scenario questionnaire thatincluded demographic questions and questions related to par-ticipants’ subjective levels of fatigue and stress, and familiaritywith the type of scenario encountered. The eye-tracker wasthen demonstrated to each participant, and the device fittedand calibrated using the recommended five-point calibrationprocedure.

Each participant took part in two consecutive scenarios sepa-rated by a 5 min interval. They waited outside the cubicle as thescenario was set up. The two scenarios were each of 3 min and 30 sduration and involved acute desaturation in a baby with bron-chiolitis, due to either coughing (Scenario A) or a seizure/apnoea(Scenario B). A nurse was present in each scenario and briefed theclinician on the condition of the child over an equivalent periodof time. The condition recovered spontaneously regardless of thetreatment given.

Prior to each scenario, participants were reminded that theyshould regard the simulator as a real patient and that their individ-ual performance was not being reported. The scenario began withthe participant called to the bedspace by the confederate bedsidenurse who introduced the scenario with a pre-scripted state-ment and a series of responses, and remained present throughouteach scenario. Three researchers were also present in the cubi-cle during the study to monitor the eye-tracker, video-recordingand simulator. All remained silent and out of view during thescenarios.

The eye-tracker automatically recorded eye movement data.Data for each participant were collated, including the number offixations, the duration of fixations in milliseconds (dwell time),the number of blinks, the number of saccades, and the range ofgaze. Video footage, taken from the perspective of participants,was also recorded throughout the tasks. The software packageBeGazeTM was used to align longitudinal data with video footagefor the purposes of analysis. Video footage was analyzed frame byframe to identify AOI. There were six AOI defined in the visualscene, namely the nurse, the monitor, the manikin’s head and face,the manikin’s torso, the manikin’s limbs, and the wall on whichthe equipment and oxygen outlets were located.

RESULTSDATA REDUCTIONTo derive information on the process of visual information acqui-sition during initial clinical assessment, the video analysis waslimited to the first 90 s of the scenario. Eye-tracking data for one

expert and three competent non-experts were excluded from fur-ther analysis due to failed eye-tracking calibration. There was noairway opening, bag and mask support, or cardiac compressioninitiated by participants during the period of analysis.

DESCRIPTIVE STATISTICSDescriptive statistics were generated for each of the dependentvariables. Across the participants and the scenarios, the mean dwelltime was 448.16 ms, SD = 9.75. The mean dwell times for each ofthe AOI is summarized in Table 1.

FIXATIONS AND SACCADESThree independent, mixed between-within analyses of variancewere undertaken to establish whether a relationship existedbetween participants’ level of expertise, the nature of the scenario,and eye tracking behavior, including the frequency of fixations andsaccades and the mean duration of fixations (dwell time).

No statistically significant differences were evident betweenexperts and competent non-experts in the frequency of fixations,F(1,10) = 1.97, p = 0.19, η2 = 0.17, or saccades, F(1,10) = 4.00,p = 0.07, η2 = 0.29. Similarly, eye gaze data were not signifi-cantly different between the scenarios in the frequency of fixations,F(1,10) = 3.89, p = 0.07, η2 = 0.28, the frequency of sac-cades, F(1,10) = 2.46, p = 0.15, η2 = 0.20, or the mean dwelltime, F(1,10) = 0.49, p = 0.50, η2 = 0.04.

Differences were, however, evident between experts and com-petent non-experts in the mean dwell time, F(1,10) = 6.48,p = 0.03, η2 = 0.39, with experts’ mean dwell time,X = 472.36 ms, SD = 14.89, greater than non-experts,X = 423.36 ms, SD = 12.58. No significant interaction wasevident between expertise and scenario for the frequency of fix-ations, F(1,10) = 0.01, p = 0.91, η2 < 0.01, the frequency ofsaccades, F(1,10) = 0.04, p = 0.85, η2 < 0.01, nor the mean dwelltime, F(1,10) = 0.58, p = 0.46, η2 = 0.05.

AOI DWELL TIME ANALYSISFor the “head and face” and “nurse” signature features, a 2 (exper-tise) × 2 (scenario) mixed-between ANOVA was used to testwhether differences existed in the overall dwell time for expertsand competent non-experts during the 90 s, initial assessment ofthe patient. It was assumed that differences in dwell time wouldreflect differences in the relative attention to features associatedwith the particular scenario. Consistent with expectations, theresults revealed a significant difference between competent non-experts and experts’ mean overall dwell time for “head and face,”

Table 1 | Mean dwell time (ms) by area of interest.

Area of interest Mean Std. error

Head 7183.407 1682.979

Torso 10203.959 2537.263

Limbs 374.330 143.209

Monitor 9628.809 1764.095

Nurse 4396.192 1633.480

Wall 1308.063 538.293





F(1,12) = 6.16, p = 0.03, η2 = 0.34, whereby experts recordedsignificantly greater dwell time within the AOI, X = 11358.83 ms,SD = 2698.77, in comparison to competent non-experts, X= 3007.98 ms, SD = 2011.55. A significant difference was alsoevident in the mean dwell time for the “nurse,” F(1,12) = 5.89,p = 0.03, η2 = 0.34. However, in this case, experts recorded asignificantly lower dwell time within the AOI, X = 433.63 ms,SD = 2619.36, in comparison to competent non-experts, X= 8358.76 ms, SD = 1952.38 (see Figure 2).

In the context of scenarios, a significant main effect was evidentfor the “head and face” AOI, F(1,12) = 5.69, p = 0.03, η2 = 0.32,whereby participants recorded a greater overall dwell time for thisAOI during the seizure scenario, X = 9438.06 ms, SD = 2452.35,in comparison to the coughing scenario, X = 4928.75 ms,SD = 1199.43. This suggests that, as a cohort, both experts andcompetent non-experts responded to the differences between thescenarios by changing their pattern of information acquisitionin relation to the head and face. However, there was no changeevident in the overall dwell time on the nurse.

At a more detailed level, an expertise by scenario interactionwas evident for the mean overall dwell time on the “head andface,” F(1,12) = 4.82, p = 0.04, η2 = 0.29. An inspection of themeans indicated that, where there was relatively little differencebetween the mean dwell times for competent non-experts acrossthe two scenarios, a difference was evident for experts with themean dwell time greater during the seizure scenario than duringthe coughing scenario (see Figure 3). In combination, these resultssuggest that, although competent non-experts may recognize therelative importance of signature features during different diag-nostic scenarios, their pattern of interaction with these featuresremains relatively consistent. This contrasts with expert clinicianswho appear to alter both the overall time that they devote to theacquisition of information from signature features, together withthe pattern of acquisition.

GENERAL DISCUSSIONThe aim of this study was to examine the information acquisitionstrategies employed by expert and competent non-expert inten-sive care physicians during two diagnostic scenarios that differed in

FIGURE 2 | Mean total dwell time (ms) and standard error on the

“confederate nurse” for experts and competent non-experts,

distributed across the two scenarios.

FIGURE 3 | Mean total dwell time (ms) and standard error on the

“head and face” for experts and competent non-experts, distributed

across the two scenarios.

their immediate features, but incorporated a similar intrinsic etiol-ogy. It was anticipated that where competent non-experts wouldadopt a relatively consistent pattern of information acquisitionacross the scenarios, experts would vary their approach consistentwith the differences in the immediate features that were presented.The results revealed differences in overall mean fixation timesbetween experts and competent non-experts, with the formermaintaining visual gaze on AOIs for significantly longer periods.Further, experts spent significantly more dwell time within the“head and face” AOI and significantly less time within the “nurse”AOI in comparison to competent non-experts.

The differential performance amongst experts and competentnon-experts during information acquisition is consistent with theproposition that experts and competent non-experts differ in theirrepertoire of cues in memory (Ericsson and Kintsch, 1995; Jaro-dzka et al., 2010). Experts attended to the immediate visual featuresassociated with the patient (“head and face”), where competentnon-experts tended to spend a greater proportion of the time fix-ated on the confederate nurse. Since this occurred independentof scenarios, it might be surmised that experts were integratingthe auditory information being delivered by the confederate nursewith the visual information that was evident from the head andface of the manikin. It also implies that the“head and face”embod-ied a greater level of diagnostic information than was availablefrom the nurse in isolation.

Despite the fact that, overall, experts tended to spend relativelymore time than competent non-experts attending to the “headand face,” differences were evident in the mean dwell times acrossthe scenarios. For competent non-experts, the relative empha-sis on the “head and face” and “nurse” did not change with thechange in scenario, suggesting that non-experts did not necessar-ily discriminate between the scenarios based on the immediatefeatures.

As hypothesized, expert physicians recorded greater mean dwelltimes on the “head and face” during the seizure scenario, thanduring the coughing scenario. This reflects the potentially greaterutility of the “head and face” in yielding diagnostic informationduring the coughing scenario. The dwell time for the “confeder-ate nurse” did not change statistically for experts, possibly dueto a restriction of range associated with the mean dwell times.In combination, the outcomes suggest that overall, experts spent





more time examining cues arising from the “head and face” ofthe patient, but that differences in the immediate features wereassociated with differences in the time spent examining the cues.

Although the results confirm that experts differ from competentnon-experts in their acquisition of information during diagnosticscenarios, it also suggests that their attention toward features in theenvironment is influenced by the interaction between their task-related experience and the immediate features that are present.For example, it is possible that, for competent non-experts, thesituation was relatively unfamiliar and, therefore, they were seek-ing information that would correspond to a relatively limitednumber of patterns in memory. Since the “confederate nurse”was delivering an initial assessment of the symptoms, and mayhave experienced the event previously, directing attention towardthe nurse represents a reasonable strategy, where a scenario isunfamiliar.

By contrast, experts possess a repertoire of cues in memoryand therefore, are drawn toward features that are implicitly diag-nostic of a particular condition (Croskerry, 2009b). The relativeproportion of attention that is directed toward signature featuresis consistent with a bottom-up recognition process, whereby theenvironmental features trigger associations in memory, and a serialprocess of pattern matching is undertaken until a corresponding(or near to corresponding) pattern is identified (Patel and Groen,1986; Klein, 2008).

At an applied level, the results suggest differences in the diag-nostic strategies employed by experts and competent non-experts,and there are implications for training. For example, the fact thatcompetent non-experts tended to attend to the nurse, suggeststhat they lacked a repertoire of cues in memory, necessary to rec-ognize and adapt to the differences in the immediate features thatwere presented. This was not the case for experts who were ableto identify the immediate features associated with the differentscenarios and respond appropriately. One approach to the devel-opment of cues in memory involves cue-based training in whichlearners participate in a series of scenarios, the aim of which isto establish the relationship between features and events/objectsin memory in the form of cues (Wiggins and O’Hare, 2003). Theutility of cue-based training has been established in other domains(e.g., Auditors), and may be appropriate for diagnostic tasks in themedical context (Earley, 2001).

LIMITATIONS AND FUTURE RESEARCHWhile a weakness of this study is the relatively limited numberof participants, the fact that differences were observed betweenexperts and competent non-experts in relation to dwell timespoints toward the underlying power of the effects that wereobserved. Moreover, the study demonstrated that, in naturalisticenvironments, where the number of features available is relativelyconstrained and where the least experienced operators are in factcompetent, differences in information acquisition were evident.

Since the focus of this study was information acquisition behav-ior during the initial assessment of a potentially deterioratingpatient, the complexity associated with therapeutic interventionswas excluded. Nevertheless, it is possible that a more extendedobservation may have revealed new information in the attentionto cues, and the interactions with auditory and tactile stimuli.

While these stimuli, were experimentally controlled in the presentstudy, future research should be directed toward examining therelative impact of communication, and the social processes thatare engaged by different groups of physicians. This builds on thebaseline data that has been established in the present study andcontributes to a broader understanding of non-visual stimuli orcues, and the role of team and social interactions in the recognitionof the deteriorating child by skilled clinicians.

CONCLUSIONThis study demonstrated differences in the information acqui-sition behavior of experts and competent non-experts duringassessments of a deteriorating child during two in situ simulations.Compared to competent non-experts, experts attended to specificvisual features for longer periods, and exhibited longer dwell timeson the manikin’s “head and face,” particularly during the seizurescenario. By contrast, competent non-experts displayed longerdwell times on the “confederate nurse.” These results were inter-preted as evidence of differences between experts and competentnon-expert physicians’ diagnostic cues in memory. The methodol-ogy offers a potential framework to develop behavioral standardsof cue acquisition and utilization that could ultimately be usedfor the assessment of the diagnostic performance of physicians,particularly in time-constrained situations.

REFERENCESAbernethy, B. (1990). Anticipation in squash: differences in advance cue uti-

lization between expert and novice players. J. Sports Sci. 8, 17–34. doi:10.1080/02640419008732128

Anderson, J. (1982). Acquisition of cognitive skill. Psychol. Rev. 89, 369–406. doi:10.1037/0033-295X.89.4.369

Bilalic, M., McLeod, P., and Gobet, F. (2010). The mechanism of the Einstellung (Set)effect: a pervasive source of cognitive bias. Curr. Dir. Psychol. Sci. 19, 111–115.doi: 10.1177/0963721410363571

Brunswik, E. (1955). Representative design and probabilistic theory in a functionalpsychology. Psychol. Rev. 62, 193–217. doi: 10.1037/h0047470

Croskerry, P. (2003). The importance of cognitive errors in diagnosis and strategiesto minimize them. Acad. Med. 78, 775–780. doi: 10.1097/00001888-200308000-00003

Croskerry, P. (2009a). A universal model of diagnostic reasoning. Acad. Med. 84,1022–1028. doi: 10.1097/ACM.0b013e3181ace703

Croskerry, P. (2009b). Clinical cognition and diagnostic error: applications of a dualprocess model of reasoning. Adv. Health Sci. Educ. Theory Pract. 14, 27–35. doi:10.1007/s10459-009-9182-2

Earley, C. E. (2001). Knowledge acquisition in auditing: training novice auditors torecognize cue relationships in real estate valuation. Account. Rev. 76, 81–97. doi:10.2308/accr.2001.76.1.81

Ely, J. W., Graber, M. L., and Croskerry, P. (2011). Checklists to reduce diagnosticerrors. Acad. Med. 86, 307–313. doi: 10.1097/ACM.0b013e31820824cd

Ericsson, K. A., and Kintsch, W. (1995). Long-term working memory. Psychol. Rev.102, 211. doi: 10.1037/0033-295X.102.2.211

Hamm, R. M. (2014). Figure and ground in physician misdiagnosis: metacognitionand diagnostic norms. Diagnosis 1, 29–33. doi: 10.1515/dx-2013-0019

Jarodzka, H., Scheiter, K., Gerjets, P., and Van Gog, T. (2010). In the eyes of thebeholder: how experts and novices interpret dynamic stimuli. Learn. Instr. 20,146–154. doi: 10.1016/j.learninstruc.2009.02.019

Kahneman, D., and Klein, G. (2009). Conditions for intuitive expertise: a failure todisagree. Am. Psychol. 64, 515–526. doi: 10.1037/a0016755

Khader, P. H., Pachur, T., Meier, S., Bien, S., Jost, K., and Rösler, F. (2011).Memory-based decision-making with heuristics: evidence for a controlled acti-vation of memory representations. J. Cogn. Neurosci. 23, 3540–3554. doi:10.1162/jocn_a_00059

Klein, G. (2008). Naturalistic decision making. Hum. Factors 50, 456–460. doi:10.1518/001872008X288385





Loveday, T., and Wiggins, M. W. (2014). Cue utilization and broad indica-tors of workplace expertise. J. Cogn. Eng. Decis. Mak. 8, 98–113. doi:10.1177/1555343413497019

Loveday, T., Wiggins, M. W., Harris, J., Smith, N., and O’Hare, D. (2013a).An objective approach to identifying diagnostic expertise amongst powersystem controllers. Hum. Factors 55, 90–107. doi: 10.1177/0018720812450911

Loveday, T., Wiggins, M. W., Searle, B. J., Festa, M., and Schell, D. (2013b). Thecapability of static and dynamic features to distinguish competent from gen-uinely expert practitioners in pediatric diagnosis. Hum. Factors 55, 125–137. doi:10.1177/0018720812448475

Mann, D. T., Williams, A. M., Ward, P., and Janelle, C. M. (2007). Perceptual-cognitive expertise in sport: a meta-analysis. J. Sport Exerc. Psychol. 29, 457.

Marewski, J. N., and Gigerenzer, G. (2012). Heuristic decision making in medicine.Dialogues Clin. Neurosci. 14, 77–89.

Morrison, B., Wiggins, M. W., Bond, N., and Tyler, M. (2013). Measuring cuestrength as a means of identifying an inventory of expert offender profiling cues.J. Cogn. Eng. Decis. Mak. 7, 211–226. doi: 10.1177/1555343412459192

Norman, G., Sherbino, J., Dore, K., Wood, T., Young, M., Gaissmaier, W., et al.(2014). The etiology of diagnostic errors: a controlled trial of system 1 versussystem 2 reasoning. Acad. Med. 89, 277–284. doi: 10.1097/ACM.0000000000000105

Patel, V. L., and Groen, G. J. (1986). Knowledge based solution strategies in medicalreasoning. Cogn. Sci. 10, 91–116. doi: 10.1207/s15516709cog1001_4

Patel, V. L., and Groen, G. J. (1991). “The general and specific nature of medicalexpertise: a critical look,” in Toward a General Theory of Expertise, eds K. A.Ericsson and J. Smith (Cambridge: Cambridge University Press), 93–125.

Pham, J. C., Aswani, M. S., Rosen, M., Lee, H., Huddle, M., Weeks, K., et al. (2012).Reducing medical errors and adverse events. Annu. Rev. Med. 63, 447–463. doi:10.1146/annurev-med-061410-121352

Shanteau, J. (1988). Psychological characteristics and strategies of expert decision-makers. Acta Psychol. 68, 203–215. doi: 10.1016/0001-6918(88)90056-X

Simon, H. A. (1972). Theories of bounded rationality. Decis. Organ. 1,161–176.

Stokes, A. F., Kemper, K., and Kite, K. (1997). “Aeronautical decision-making, cuerecognition, and expertise under time pressure,” in Naturalistic Decision-Making,eds C. Zsambok and G. Klein (Mahwah, NJ: Lawrence Erlbaum), 183–196.

Stolper, E., Van de Wiel, M., Van Royen, P., Van Bokhoven, M., Van der Weijden,T., and Dinant, G. J. (2011). Gut feelings as a third track in general practitioners’diagnostic reasoning. J. Gen. Intern. Med. 26, 197–203. doi: 10.1007/s11606-010-1524-5

Sweller, J. (1988). Cognitive load during problem solving: effects on learning. Cogn.Sci. 12, 257–285. doi: 10.1207/s15516709cog1202_4

Wiggins, M., and O’Hare, D. (2003). Weatherwise: evaluation of a cue-based trainingapproach for the recognition of deteriorating weather conditions during flight.Hum. Factors 45, 337–345. doi: 10.1518/hfes.45.2.337.27246

Wiggins, M. W. (2006). “Cue-based processing and human performance,” in Ency-clopedia of Ergonomics and Human Factors, 2nd Edn, ed. W. Karwowski (London:Taylor and Francis), 641–645.

Wiggins, M. W. (2012). The role of cue utilisation and adaptive interface designin the management of skilled performance in operations control. Theor. IssuesErgon. Sci. 15, 283–292. doi: 10.1080/1463922x.2012.724725

Wiggins, M. W., Azar, D., Hawken, J., Loveday, T., and Newman, D. (2014). Cue-utilisation typologies and pilots’pre-flight and in-flight weather decision-making.Saf. Sci. 65, 118–124. doi: 10.1016/j.ssci.2014.01.006

Wiggins, M. W., and O’Hare, D. (1995). Expertise in aeronautical weather-relateddecision-making: a cross-sectional analysis of general aviation pilots. J. Exp.Psychol. Appl. 1, 305–320. doi: 10.1037/1076-898X.1.4.305

Conflict of Interest Statement: The authors declare that the research was conductedin the absence of any commercial or financial relationships that could be construedas a potential conflict of interest.

Received: 08 May 2014; accepted: 08 August 2014; published online: 26 August 2014.Citation: McCormack C, Wiggins MW, Loveday T and Festa M (2014) Expert andcompetent non-expert visual cues during simulated diagnosis in intensive care. Front.Psychol. 5:949. doi: 10.3389/fpsyg.2014.00949This article was submitted to Cognition, a section of the journal Frontiers in Psychology.Copyright © 2014 McCormack, Wiggins, Loveday and Festa. This is an open-access arti-cle distributed under the terms of the Creative Commons Attribution License (CC BY).The use, distribution or reproduction in other forums is permitted, provided the originalauthor(s) or licensor are credited and that the original publication in this journal is cited,in accordance with accepted academic practice. No use, distribution or reproduction ispermitted which does not comply with these terms.






ORIGINAL RESEARCH ARTICLEpublished: 10 June 2014

doi: 10.3389/fpsyg.2014.00556

Spontaneously spotting and applying shortcuts inarithmetic—a primary school perspective on expertiseClaudia Godau1,2*, Hilde Haider3, Sonja Hansen3, Torsten Schubert1,2, Peter A. Frensch1,2 and

Robert Gaschler2,4

1 Department of Psychology, Humboldt-Universität zu Berlin, Berlin, Germany2 Cluster of Excellence: Image Knowledge Gestaltung, an Interdisciplinary Laboratory, Berlin, Germany3 Department of Psychology, Universität Köln, Köln, Germany4 Department of Psychology, Universität Koblenz-Landau, Landau, Germany

Edited by:


Reviewed by:

Ann Dowker, University of Oxford,UKLieven Verschaffel, University ofLeuven, Belgium

*Correspondence:

Claudia Godau, Institut fürPsychologie, Fakultät fürLebenswissenschaften, HumboldtUniversität zu Berlin, Unter denLinden 6, 10099 Berlin, Germanye-mail: [email protected]

One crucial feature of expertise is the ability to spontaneously recognize where andwhen knowledge can be applied to simplify task processing. Mental arithmetic is onedomain in which people should start to develop such expert knowledge in primaryschool by integrating conceptual knowledge about mathematical principles and proceduralknowledge about shortcuts. If successful, knowledge integration should lead to transferbetween procedurally different shortcuts that are based on the same mathematicalprinciple and therefore likely are both associated to the respective conceptual knowledge.Taking commutativity principle as a model case, we tested this conjecture in twoexperiments with primary school children. In Experiment 1, we obtained eye trackingdata suggesting that students indeed engaged in search processes when confrontedwith mental arithmetic problems to which a formerly feasible shortcut no longerapplied. In Experiment 2, children who were first provided material allowing for onecommutativity-based shortcut later profited from material allowing for a different shortcutbased on the same principle. This was not the case for a control group, who had firstworked on material that allowed for a shortcut not based on commutativity. The resultssuggest that spontaneous shortcut usage triggers knowledge about different shortcutsbased on the same principle. This is in line with the notion of adaptive expertise linkingconceptual and procedural knowledge.

Keywords: expertise, numerical cognition, arithmetic, commutativity, spontaneous strategy application

INTRODUCTIONExpertise has various manifestations and could be defined asconsistently superior performance within a specific domain rel-ative to novices and relative to other domains (Ericsson andLehmann, 1996). The development of expertise in real-worlddomains involves a complex interplay of changes in perception,categorization, memory, problem solving, coordination, skilledaction, and other components of human cognition (Palmeri et al.,2004). Expert’s flexibility has been frequently discussed and thereexist two contradictory perspectives. Research on creativity andskill acquisition has been used to illustrate that more knowl-edge can make one less flexible (i.e., Luchins, 1942; Logan, 1988).However, research on expertise suggested that experts are moreflexible and creative in their thought patterns (see summary inBilalic et al., 2008a). Both options might be possible depending onthe expertise level and the problem difficulty. Investigating chessexperts Bilalic et al. (2008a) found that “super experts” were flex-ible and find the optimal solution first or at least find it quicklyafter perceiving a salient but non-optimal solution.

Here, we focus in the domain of mathematics on spon-taneously spotting and applying shortcuts in arithmetic andwhether with further experience students become increasinglyable to generate rapid adequate actions with less and less effort

(Ericsson, 2008). Mathematic students used significantly largernumbers of appropriate strategies than adults with less expertise(Dowker et al., 1996). Experts have to be able to recognize spon-taneously and without instruction that a specific element of theirknowledgebase can be applied in a specific situation. It would notsuffice if they possessed elaborate conceptual knowledge as wellas procedures to apply it, but needed to wait for someone to tellthem that the knowledge can be applied in the given situation.This someone would rarely drop by.

In recent years, research in primary school arithmetic hasstarted to tackle this issue for a domain in which everyoneshould acquire elaborate knowledge. Learning about mathemati-cal principles and procedures should lead to knowledge that canbe applied across a wide range of situations (e.g., Hatano andOura, 2003). Given the role of self-guided learning and perfor-mance in the development of mathematical abilities and concepts,recent studies have focused on the question how and when chil-dren spontaneously recognize that an everyday situation can betackled by mathematical thinking (Hannula and Lehtinen, 2005;Hannula et al., 2010; McMullen et al., 2011). Furthermore, chil-dren should develop the skills necessary to flexibly spot andapply shortcut strategies spontaneously. It is not sufficient if theycan apply a shortcut when explicitly told to do so. Adaptive

www.frontiersin.org June 2014 | Volume 5 | Article 556 | 24













Godau et al. From shortcut to shortcut

expertise (Verschaffel et al., 2009) includes to autonomously reg-ulate whether (a) to solve an arithmetic problem in a standardway or to (b) search for / apply a shortcut.

Taking the commutativity principle as a model case, pastresearch has explored how children spontaneously spot and applyshortcuts that allow saving effort in addition problems by flexiblychanging the order of addends. Wealth of research has shown thatchildren have at least some understanding of the concept of com-mutativity before entering school (Baroody and Gannon, 1984;Resnick, 1992; Cowan and Renton, 1996; Wilkins et al., 2001;Canobi et al., 2003). After interviewing elementary school chil-dren how they solved problems with two addends, (Baroody et al.,1983) report an extensive use of commutativity. During develop-ment children increasingly integrate conceptual knowledge aboutmathematical principles and procedural knowledge about short-cuts (Haider et al., 2014). Knowledge integration should lead totransfer between procedurally different shortcuts that are basedon the same mathematical principle and therefore likely bothassociated to the respective conceptual knowledge. In a first step,(Gaschler et al., 2013) provided a correlative study to explorethis idea. They assessed spontaneous usage of two procedurallydifferent shortcuts that are both based on the commutativityprinciple in children of different age. While shortcut usage wasobserved from second grade onwards, correlations between theusage of the two different shortcuts only emerged by grade four.In the current study we aimed at moving beyond correlationaldata. We tested whether being exposed to one commutativity-based shortcut helps to spot and apply a different shortcut optionbased on the same mathematical principle. Note that in a par-allel line of research, we have observed that instructions do notseem to do the job. Instructing children to use one specific short-cut does hinder rather than assist them in spontaneously spottingand applying a different shortcut based on the same mathemat-ical principle later on (Godau et al., submitted). Instructionsabout specific procedures might corrupt flexibility in shortcutusage (cf. ErEl and Meiran, 2011). Even when participants knewthat a formerly instructed rule would no longer apply, theyfound it difficult to search for different shortcut options (see alsoBilalic et al., 2008a,b; Bilalic and McLeod, 2014). Therefore, inthe current work we focused on spontaneous use of the strate-gies. We explored whether it is possible to foster the discoveryand application of shortcut strategies by transfer between dif-ferent non-instructed shortcut strategies that are based on thesame mathematical principle. Note that according to Baroody andGannon (1984) understanding of commutativity was not evidentin all those who invented shortcuts, but in all those who compre-hend addition as a binary rather than as a unary operation. Theunary view would suggest that one number is added to another,rather than that they are added together.

Specifically, the commutativity principle enables students toflexibly change the order of addends within a problem. Forinstance, given the problem 4 + 7 + 6, it might be easier to cal-culate (6 + 4) +7 (6 + 4 adds up to 10 which makes it easy tofinally add 7, i.e., “Ten-strategy”). One can also use commutativ-ity across problems. If, for instance, a student receives the problem8 + 5 + 7 =?, and then 5 + 7 + 8 =?, he/she can refrain fromcalculating the second problem presupposed he / she recognizes

the applicability of the commutativity principle (i.e., “addends-compare strategy”). Three-addends problems were used, becausewe wanted to investigate usage of the commutativity princi-ple with unfamiliar problems. It is debatable if three-addendsproblems imply only the commutativity principle or additionallyalso the associativity principle. Associativity is the property thatproblems in which terms are decomposed, and recombined in dif-ferent ways, have the same answer [(a + b) + c = a + (b + c)]. Inthe problems we used, children have to change the order of theaddends [a + b + c = (a + c) + b], because otherwise it is notpossible to add a + c first. Commutativity justifies changing theorder or sequence of the operands within an expression whileassociativity does not.

In Experiment 1, we used eye tracking to explore how chil-dren search and apply different commutativity-based shortcuts.Verschaffel et al. (1994) presented third-graders with three-addends problems and assessed eye movements combined withverbal report and found that in 71% of all possible cases com-mutativity was used. We used a different approach, as we ratherwere interested in whether children spontaneously start searchprocesses when, after a change in the material one shortcut optionis no longer present. The findings suggested that being offered anopportunity to apply one commutativity-based shortcut can helpto search for and apply a different shortcut based on the sameprinciple when the first one is no longer feasible. In Experiment 2,we explored whether transfer from shortcut to shortcut might beconcept specific: on the one hand, it seems plausible that short-cuts based on the same mathematical principle trigger each otherbecause they are linked to one-another directly or indirectly (asthey are both linked to the common conceptual knowledge). Thisperspective is in line with research suggesting that mathemati-cal knowledge develops in an iterative fashion, with conceptualchange influencing procedural change and vice versa (Byrnes andWasik, 1991; Hiebert and Wearne, 1996; Rittle-Johnson et al.,2001; Waldmann, 2006). For instance, Canobi (2009) showedthat children’s conceptual advances were predicted by their ini-tial procedural skills. On the other hand, transfer from shortcutto shortcut might occur place for motivational reasons unrelatedto the specific shortcut and underlying mathematical principle.After having experienced that task processing can be simplifiedby a shortcut, one might be more apt to search for and apply anyshortcut, as one has learned that attractive shortcut options doseem to exist in the material provided.

EXPERIMENT 1In Experiment 1, we used eye tracking in order to explore the fix-ation patterns reflecting the usage of shortcut strategies. We werefurthermore interested in how fixation patterns reflect how peo-ple accommodate to being presented with new sets of arithmeticproblems within which the previously feasible shortcut no longerapplies (but instead a different shortcut). To this end, children atfirst had to solve problems that could be facilitated by the ten-strategy (of three addends, the first and the last add up to 10).After that, they were presented with problems that allowed forthe use of the addends-compare strategy (some problems con-tained the same addends as their precursor in different order).Both strategies are based on the commutativity principle.

Frontiers in Psychology | Cognition June 2014 | Volume 5 | Article 556 | 25





METHOD EXPERIMENT 1ParticipantsTwenty children participated in Experiment 1 (mean age 8.6years). They were tested individually in a laboratory at Humboldt-Universität, Berlin.

Procedure and MaterialsResearch procedures of these experiments were approved in apeer review process for applying for public funding (GermanResearch Foundation, DFG) and were completed with approval ofthe Institutional Review Board of the Department of Psychologyat Humboldt-Universität, Berlin. Students were informed aboutthe content of the study and that data analysis would preserveanonymity. We ensured written informed consent of the parents.Children were than tested individually with a 250 Hz video-basedeye tracker (SMI RED 250). Packages of six problems in blackon a gray background were shown on a 22 TFT monitor, withthe student sitting at approximately 50 cm distance. Digits wereapproximately 0.5 cm wide and 1 cm tall.

Children started with a five-point calibration. Afterwards theexperimenter showed a single example problem and explainedthat the children should utter the result as quickly and as accu-rately as possible. Children started the main part by working ontwo screens with six ten-strategy problems each (first and lastaddend add up to 10). They then completed two screens withaddends-compare problems intermixed with baseline problems.Two of six problems per screen contained identical addends indifferent order as the preceding problem (problems listed in theSupplementary materials). Each problem was presented in oneline and consisted of three different addends between 2 and 9(maximum result was 24; 0 and 1 were excluded as addends). Webalanced problem size between the addends-compare problemsand the baseline problems so that they were equally difficult forchildren unless they used the shortcut (for more details Gaschleret al., 2013; Haider et al., 2014).

Children were presented the first screen (of two) with six ten-strategy problems. The experimenter moved the cursor to theright of the equal sign of the first problem and waited for ananswer. The answer was immediately entered as the time log ofthe first key press served to determine the calculation time asthe span from the cursor allocation to the first (i.e., two-digit

results) key press of entering the result for the current problem.After entering the answer, the experimenter moved the cursor tothe next problem. The entered results remained visible on thescreen while working on the remaining of the six problems ofthe package. This was especially important for the work on thetwo screens with addends-compare problems later on. If they hadspotted that the addends of a problem were the commuted versionof the preceding problem, that way they were provided with theopportunity to access the solution they had given on the previousproblem.

RESULTSThe computerized assessment allowed to track solution timeson the level of single problems. As previously mentioned, stu-dents calculated 12 ten-strategy problems (Screen 1 and 2) andafterwards worked on yet another 12 problems, four of themallowed for the addends-compare strategy (Screen 3 and 4).Figure 1 shows the mean solution times per problem for eachscreen. Students were faster on addends-compare problems ascompared to baseline problems. A 2 (screen: first vs. second) × 2(problem type: addends-compare problem vs. baseline problem)ANOVA with solution times as dependent variable revealed a sig-nificant main effect of problem type, [F(1, 19) = 7.46, p = 0.01,η2

p = 0.28]. Neither the main effect of screen, [F(1, 19) = 1.67,

p = 0.21, η2p = 0.08], nor the interaction effect were significant,

[F(1, 19) = 0.72, p = 0.41, η2p = 0.04]. We did not find signifi-

cant effects when repeating the above analyses with error rate asdependent variable (see Supplementary materials).

The analysis of the eye tracking data suggests that the ten-strategy and the addends-compare strategy can be identified byspecific fixation patterns. Using the ten-strategy, adding the firstand last addend first to receive the result ten, should be fast andnecessitates little fixation time on the outer numbers. Addingthe middle number afterwards and uttering the result mighttherefore result in more fixation time on the middle number rel-ative to the other numbers. Figure 2 suggests that the percentfixations falling on the middle vs. outer numbers of the three-addends problems are distributed in line with this reasoning. Thepercentage of fixations on the middle number increased fromthe first to the second screen of the ten-strategy problems, asstudents presumably discovered the structure of the problems.

4

6

8

10

12

14

Ten-strategy screen 1

Ten-strategy screen 2

Addends-compare strategy

screen 1

Addends-compare strategy

screen 2

Solv

ing

time

per

arith

met

ic

prob

lem

in se

cond

s

Baseline

Shortcut-strategy (ten-/ addends-compare)

FIGURE 1 | Mean calculation time per arithmetic problem in Experiment 1. Error bars indicate the standard error of the mean.





20

22

24

26

28

30

32

Ten-strategy screen 1 Ten-strategy screen 2 Addends-compare strategy screen 1

Addends-compare strategy screen 2

% fi

xatio

ns o

n th

e m

iddl

e di

git

FIGURE 2 | Percent fixation frequency on second addend for

ten-strategy and addends-compare strategy booklets. The error bardisplays the standard error of the mean.

When the ten-strategy could no longer be used (first screen withaddends-compare problems), the percent fixations on the middledigit were low again. Surprisingly, it increased on the secondscreen with addends-compare problems. A 2 (screen: first vs. sec-ond) × 2 (ten-strategy problems vs. addends-compare problems)ANOVA with percentage of fixations falling on the middle num-ber as dependent variable revealed a significant main effect forstrategy [F(1, 17) = 6.02, p < 0.05, η2

p = 0.26]. Children fixatedthe middle digit more in problems, in which the ten-strategycould be used compared to problems on the addends-comparescreens. There was also a significant main effect of screen,[F(1, 17) = 7.91, p = 0.01, η2

p = 0.32], but no interaction, F < 1.In the ten-strategy problems, addends should be checked

within a line in order to identify shortcut options. In contrast,for the addends-compare strategy, it is necessary to compare theaddends between the lines. Children should thus not only fix-ate the addition problem they are currently solving but also theprevious one or the subsequent one in order to check whethera set of addends repeats. Figure 3 presents the mean differencesbetween (a) line fixated and (b) line of current problem. If, forinstance, a student during solving a problem was fixating back onthe problem in the line before, this would lead to a value of −1 forthis particular fixation. While the majority of fixations were onthe line of the current problem, some fixations were directed atprevious (negative difference) or subsequent (positive difference)problems. We focused on comparing the above index of fixationposition between the addends-compare problems and their pre-ceding problems. Thus, addends are identical and only differ inorder. We found a significant difference in the index of fixationposition for these problems. In line with our assumption, studentswere fixating ahead on problems preceding the addends-compareproblems and fixating back, once a set with identical addends wasdiscovered, [t(18) = 5.44, p < 0.001].

In addition to identifying eye movement patterns that are spe-cific for the shortcuts we found a significant correlation betweenthe increase of the fixation on the middle digit in the ten-strategy problems (Screen 2—Screen 1) and the time benefit onaddends-compare strategy problems r = 0.49, p = 0.05. Thus,

increased usage of the commutativity-based shortcut offered onScreen 1 and Screen 2 might help in spotting and applying theother commutativity-based shortcut offered on Screen 3 and 4.

DISCUSSIONProviding children with the opportunity to spontaneously (with-out instruction or other hints) use one commutativity-basedshortcut might help them to spot and apply another shortcutbased on the same mathematical principle once the first one doesno longer apply. Furthermore, the eye tracking data are in linewith the interpretation that search processes might start once oneshortcut no longer applies. We found that children in some caseschecked addends of subsequent addition problems in advance(i.e., before uttering the result to the current problem and theallocation of the cursor to the next problem). Note that thisimplies that the accuracy to attribute calculation time to specificarithmetic problems might be limited in setups in which multi-ple problems are simultaneously presented. Such arrangementsresemble work on arithmetic problems on worksheets in theschooling context. Eye tracking or reliance on aggregate measuresfrom paper-and-pencil versions might both be useful approachesto this variant of the dilemma of external vs. internal validity.

Experiment 1 provided a first hint in line with the idea thatthere might be transfer from one shortcut to another one. Thissuggests two different explanations. On the one hand, sponta-neously spotting and applying shortcuts on Screen 1 and 2 mightaffect processing of Screens 3 and 4 on a motivational route.Participants learn that shortcut options seem to exist and can beexploited. This would suggest that such transfer could take placefrom any easily identifiable shortcut to a second one. On the otherhand, transfer might involve specific mathematical knowledge. Itmight first and foremost take place between shortcuts based onthe same mathematical principle. We tried to disentangle thesetwo perspectives in Experiment 2.

EXPERIMENT 2This experiment focused on the question if the ten-strategy facil-itated the usage of the addends-compare shortcut. For this pur-pose, we compared three conditions: students in the ten-strategywarm-up condition started with the ten-strategy problems fol-lowed by problems that allowed for applying the addends-compare strategy (similar to Experiment 1). In the baselinewarm-up condition, children worked on material with no short-cut option at all before being transferred to the addends-comparebooklet. The inversion warm-up condition started with inver-sion problems (e.g., 9 + 2 − 2). Thus, a shortcut not based onthe commutativity principle was offered first. This was impor-tant in order to test whether all shortcut strategies would alterthe usage of the addends-compare shortcut simply by motivationchildren to look for shortcuts. Alternatively, it might be that onlythe ten-strategy increases the probability to spot the addends-compare strategy, as it is the only shortcut strategy, which isalso based on the commutativity principle. It is conceivable thatoffering problems with an easy-to-find shortcut option (inversionor ten-strategy) might lead students to assume that it is worth-while to search for shortcut options in general in later material.This could accordingly lead to transfer which is simply based on






-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

1

Mean difference between line

fixated and line of current problem

Before addends-compare

Addends-compare

FIGURE 3 | Mean difference between current line fixated and line of current problem. Negative values indicate fixations on preceding problems while positivevalues result from fixations on subsequent problems. Error bars indicate the 95% CI of the comparison of addends-compare problems vs. preceding problems.

the motivation to search for shortcuts. In contrast, a finding oftransfer for the ten-strategy problems but not for the inversionproblems would suggest that indeed triggering the basic principleof commutativity is important for transfer to occur.

METHOD EXPERIMENT 2ParticipantsWe tested 153 children at the end of second grade (most of themwere taught in combined classes of first and second grade) and140 children in third grade. We ensured written informed con-sent of the parents in collaboration with the schools. Either groupwas provided with advance information concerning the contentof the study (calculating mental arithmetic problems) and wasinformed that participation was voluntary. Parents and studentswere also informed that data analysis would preserve anonymity.Data were acquired in a classroom setting with paper and pencil.Gender was balanced as much as possible. Eleven children (sec-ond grade) and 20 children from the third grade were excludedby median ± 3 MADs. The MAD is a robust method to detectoutliers by using absolute deviation from the median; for furtherinformation see (Leys et al., 2013). For the descriptive data of thesample see Table 1.

Procedure and MaterialsThe arithmetic problems were the same as in Experiment 1 andare listed in the Supplementary materials. Each problem waspresented in one line and consisted of three different addendsbetween 2 and 9 (maximum result was 24; 0 and 1 wereexcluded as addends). The different types of problems werepresented as a paper pencil test in separate booklets. As depen-dent variable we measured the number of problems solved inthe booklet that allowed vs. the booklet that did not allow forthe addends-compare strategy. We took care that the amountof time provided per booklet was not sufficient to solve allproblems so that we could use number of problems solved pertime as a dependent variable (see Table 1 for time provided perbooklet).

Table 1 | Sample data and time provided per booklet in Experiment 2.

Grade Condition/ Outliers N Mean Seconds for

warm-up (female) age (SD) addends-compare

booklets

2 Ten-strategy 4 48 (25) 7.1 (0.69) 240

Baseline 1 49 (26) 7.1 (0.72)

Inversion 6 45 (25) 7.1 (0.62)

3 Ten-strategy 5 41 (24) 8.0 (0.35) 180*

Baseline 7 40 (20) 8.2 (0.71)

Inversion 8 39 (25) 7.8 (0.64)

*We started with 210 s and than reduced it after testing one group of students

in order to avoid ceiling effects.

Experimental conditions differed in the warm-up booklet.The ten-strategy warm-up started with problems in which chil-dren could use the ten-strategy. The baseline warm-up conditionsstarted with addition problems of comparable size, but that didnot include any option for applying the commutativity princi-ple to solve the problems (e.g., 4 + 3 + 5 or 7 + 6 + 2). A secondcontrol condition, the inversion warm-up condition, started withproblems that allowed for a shortcut, but, importantly, not fora commutativity-based one. Inversion problems (e.g., 9 + 2 − 2)allow refraining from calculation by comparing the numbersinvolved in the problem mixing addition and subtraction. Thus,while the ten-strategy and addends-compare strategy are bothbased on the same arithmetic principle, inversion and addends-compare are not. However, on the surface the latter two shortcutsare similar as they both enables students to avoid calculation alto-gether (in contrast, the ten-strategy does reduce instead of avoidcalculation demands).

After the warm-up phase, all children worked on five morebooklets. Starting with (1) a booklet, where the addends-comparestrategy could be used, they then were presented (2) a baselinebooklet with no shortcut opportunities, followed by (3) anotherbooklet, where the addends-compares strategy could be used.





This second addends-compare booklet was applied as we hadobtained high variability across students as well as large generalpractice effects in the first booklets in earlier work (Gaschler et al.,2013). Booklets 4 and 5 served the purpose to control whether theinduced shortcut is known and would be used (see Table 2). Thechildren in the ten-strategy warm-up condition received anotherbooklet with addition problems allowing for the ten-strategy (4)plus afterwards a matched baseline booklet (5). This was also thecase for children of the control condition with the baseline warm-up. The children of the inversion warm-up condition workedfor the second time on a booklet with inversion problems (4)followed by a matched baseline booklet (5).

Students were instructed to solve the problems as quickly andas correctly as possible. The time for each booklet was fixed andwe counted the number of problems solved and errors. Studentswere additionally informed that it would be almost impossibleto solve all problems during the period of time given for eachbooklet. As dependent measure we calculated the average time perproblem on addends-compare booklets as compared to baselinebooklets.

RESULTSAfter the short warm-up phase, children were still rather slow incalculating the first set of addends-compare booklets and betweenstudents variability was rather high (see Table 3). On closer exam-ination, we found that the practice effects were stronger than theeffect of problem type. For further analysis we focused on the sec-ond addends-compare booklet. We first analyzed the effects of ourdifferent warm-up phases on the addends-compare problems. Forcalculating the addends-compare benefit in second graders, wesubtracted for each child the average solution time per problemin Booklet 3 (addends-compare strategy) from the average time

Table 2 | The order of the booklets in Experiment 2.

Condition/

Warm-up

Booklet 1 Booklet 2 Booklet 3 Booklet 4 Booklet 5

Ten-strategyAddends-comparestrategy

Addends-comparestrategy

Ten-strategyBaseline Baseline Baseline

Inversion Inversion

per problem in Booklet 2 (baseline). The benefits are depictedin Figure 4 separately for each of the three conditions in sec-ond and third graders. In addition, Table 3 presents the averagetime per problem for every booklet for the second and thirdgrade.

For the second graders with the ten-strategy warm-up phase,we observed a significant benefit on the addends-compare strat-egy problems compared to baseline problems t(47) = 2.48, p =0.05. Second graders with the warm-up problems not allow-ing for any shortcut did not benefit from the addends-comparebooklets relative to the baseline booklets. The inversion prob-lems group also did not show such a benefit either. Thirdgraders, however, seemed to use the addends-compare strategyin every warm-up condition. Each of the three warm-up groupssignificantly benefitted from the addends-compare strategy [ten-strategy: t(40) = 2.64, p = 0.05; baseline: t(39) = 3.71, p = 0.001;inversion: t(38) = 3.79, p = 0.001]. The time used to solve theaddends-compare strategy problems was shorter than that neededto calculate the baseline problems.

We calculated a 2 (problem type: baseline vs. addends-compare booklet) × 3 (warm-up condition: ten-strategy vs.baseline vs. inversion warm-up) × 2 (grade: second vs. thirdgrade) mixed ANOVA with mean benefit time as dependentvariable. This ANOVA yield significant main effects of prob-lem type [F(1, 256) = 14.98, p < 0.001, η2

p = 0.055] and grade

[F(1, 256) = 38.44, p < 0.001, η2p = 0.131] and a significant three-

way interaction of problem type × warm-up condition × grade[F(2, 256) = 3.75, p = 0.05, η2

p = 0.028]. We found neither a sig-nificant main effect for warm-up condition, nor other interactioneffects (see Table 4). The three-way interaction suggests thatthe different warm-up phases differentially affected second andthird graders. Whereas the ten-strategy warm-up increased theprobability of applying the addends-compare strategy in secondgraders, it did not in third graders. The results suggest that short-cut to shortcut transfer specific to the underlying mathematicalprinciple was observed in second graders. Third graders, on theother hand, maybe spontaneously used the addends-compareshortcut anyways and thus did not profit from a prior task with aconceptually related shortcut.

One could argue that second graders did not show trans-fer from an inversion warm-up to addends-compare prob-lems, because they did not discover the shortcut option in the

Table 3 | Mean time per problem and standard deviation analyzed for booklet type and grade in Experiment 2.

Grade Condition Warm-up Booklet 1: Booklet 2: Booklet 3: Benefit Booklet 4: Booklet 5:

Addends-compare Baseline Addends-compare (baseline—addends- Same as Baseline (2)

strategy (1) strategy (2) compare strategy (2)) warm-up*

2 Ten-strategy 26.4 (26.8) 28.1 (34.6) 25.2 (23.0) 20.1 (12.0) 5.1 (14.3) 21.9 (22.7) 18.4 (9.2)

Baseline 23.8 (15.2) 28.3 (24.3) 22.9 (13.4) 22.6 (14.6) 0.3 (5.7) 23.0 (18.4) 22.0 (18.3)

Inversion 26.3 (29.0) 28.2 (20.4) 25.0 (19.6) 24.4 (16.6) 0.6 (10.7) 17.8 (24.9) 24.4 (21.7)

3 Ten-strategy 10.4 (3.9) 13.2 (3.3) 13.5 (3.7) 12.4 (3.6) 1.1 (2.7) 11.1 (4.6) 12.3 (5.6)

Baseline 13.9 (7.4) 14.6 (4.5) 15.8 (5.8) 13.3 (2.9) 2.4 (4.2) 13.2 (7.3) 13.4 (5.5)

Inversion 12.3 (10.3) 15.6 (8.1) 15.8 (6.3) 13.4 (4.5) 2.4 (3.9) 5.9 (5.7) 13.1 (8.2)

*See Table 2.






inversion problems. Our manipulation checks do not supportthis alternative explanation. We analyzed the Booklets 4 and 5(induction shortcut—and respective baseline). The results sug-gested that students were capable of using the inversion strategy(see Table 5). For the second graders, a 2 (Booklet 4 vs. 5) × 3(warm-up condition) ANOVA revealed a significant interactioneffect of both factors, [F(2, 139) = 3.20, p = 0.05, η2

p = 0.044]. Itdepended on the warm-up condition, whether the shortcut inBooklet 4 was used.

For the third graders we also found an interaction effectof Booklet 4 vs. 5 and warm-up condition, [F(2, 117) = 15.41;p < 0.001, η2

p = 0.208]. While there was a pronounced inversioneffect, surprisingly, neither baseline warm-up condition nor theten-strategy warm-up condition showed a ten-strategy effect inthe booklets administered at the end of the experiment. We didnot find relevant effects when repeating the above analyses witherror rate as dependent variable, but needless to say we founddifferent error rates in grade two and three (see Supplementarymaterials).

DISCUSSIONIn Experiment 2, we tested whether it is possible to make stu-dents to spot and apply a shortcut strategy by first providing

-4

-2

0

2

4

6

8

10

Ten-strategy Baseline Inversion

Add

ends

-com

pare

str

ateg

y be

nefit

in s

econ

ds

Grade 2

Grade 3

FIGURE 4 | The mean benefit in seconds of booklets allowing for the

addends-compare strategy compared to baseline booklets for the

three different warm-up conditions (ten-strategy, baseline and

inversion) for the second grade (dark gray) and the third grade (light

gray) in Experiment 2. The error bar displays the 95% confidence intervalof the comparison with zero benefit.

an easy-to-find shortcut strategy based on the same mathemat-ical principle vs. one based on a different principle. Our findingssuggest that in second graders, transfer was related to the math-ematical principle rather than to general motivational factors.There was no indication that second graders were motivated tosearch for and apply any shortcuts after being offered the firstone. If the additional conceptual link between the two differ-ent strategies is the reason for the transfer, this would supportunderstanding of adaptive expertise as the ability to apply mean-ingfully learned procedures flexibly and creatively (Hatano andOura, 2003). The inversion warm-up phase—an easy-to-findshortcut that is not based on commutativity—did not lead toincreased usage of the addends-compare strategy. While inver-sion did not promote transfer, our manipulation check suggestedthat inversion was indeed used. This is in line with Robinson andDubé (2009) who found that the inversion shortcut is easier toapply than associativity (which is similar to commutativity). Inboth studies (Robinson and Dubé, 2009; Dubé and Robinson,2010), inversion shortcut use was far more frequent than theassociativity-based strategy. Focusing on commutativity as modelcase a limitation of the experiment is that we so far only used oneshortcut not based on commutativity (i.e., inversion) in order todifferentiate between transfer effects based on motivation vs. onmathematical principles shared by subsequently offered shortcutoptions. For instance, it would be interesting to know whetherthe current setup can be turned around with inversion usage asdependent variable and commutativity vs. inversion warm-up asindependent variable (cf. Dowker, 2014). Generalizability beyondthe specific pairing of shortcuts tested here might for instancedepend upon the relative difficulty of shortcuts used as warm-upand dependent variable.

While the results suggest that second graders profited fromshortcut-to-shortcut transfer based on commutativity, thirdgraders did not seem to benefit from such extra scaffolding.Spontaneous usage of the addends-compare strategy was notimproved further by a warm-up condition with a shortcut-optionbased on the same mathematical principle. We assume that inthis age group, the concept of commutativity is more developedso that extra support is less needed. With further experience,students become increasingly able to rapidly generate adequateactions with less and less effort (Ericsson, 2008). In line withthese findings, differences between second and third graders intheir mathematical abilities are mirrored in functional changes ofthe brain. Rosenberg-Lee et al. (2011) examined the behavioral

Table 4 | Experiment 2: Results of the ANOVA problem type × grade × condition.

F p η2p

Main effect: Problem type (addends-compare strategy vs. baseline) 14.98 0.00 0.06

Grade 38.44 0.00 0.13

Warm-up condition 0.49 0.61 0.00

Inter action: Problem type (addends-compare strategy) × grade 0.00 0.96 0.00

Problem type (addends-compare strategy) × warm-up condition 1.57 0.21 0.01

Warm-up condition × grade 0.14 0.87 0.00

Problem type (addends-compare strategy) × warm-up condition × grade 3.75 0.02 0.03





Table 5 | Results of the ANOVA problem type × condition separately for grade 2 and 3.

Grade 2 Grade 3

F p η2p F p η2

p

Main effect: Problem type (addends-compare strategy) 4.92 0.03 0.03 35.04 0.00 0.23

Warm-up condition 0.23 0.80 0.00 1.95 0.15 0.03

Inter action: Problem type (addends-compare strategy) × warm-up condition 2.97 0.06 0.04 1.73 0.18 0.03

and neurodevelopmental changes between grades 2 and 3 andfound that arithmetic complexity was associated with regionsimplicated in domain-general cognitive control but also regionsfor numerical arithmetic processing. The results showed thatbrain response and connectivity relating to an arithmetic tasksignificantly change within the narrow 1-year interval.

GENERAL DISCUSSIONWe presume that one crucial feature of expertise is the abilityto spontaneously recognize where and when knowledge can beapplied to simplify task processing. In some domains, it is neces-sary for everyday life to develop this ability. Research of expertiseshowed that experts are more flexible and creative in their thoughtpattern. For instance, “super experts” were more flexible to findan optimal solution despite distraction by a non-optimal butsalient solution of a chess problem (Bilalic et al., 2008a). Playersat lower levels of expertise reported that they were looking for abetter solution, but their eye movements showed that they con-tinued to look at features related to the solution they had alreadythought of (Bilalic et al., 2008b). For expertise in object recog-nition, Harel et al. (2013) developed an interactive framework,which posits that expertise emerges from multiple interactionswithin and between the visual system and other cognitive sys-tems, such as top-down attention and conceptual memory. Theinterplay between these other, multiple cognitive processes andperception are often not consciously accessible for the expertsthemselves (Palmeri et al., 2004).

In some parts of arithmetic, procedural and conceptual knowl-edge start to develop even before primary school. In the first yearsof primary school, integration of different fragments of procedu-ral and conceptual knowledge should lead to a knowledge basethat allows to spontaneously spot and apply shortcut optionsalready in primary school. If successful, knowledge integrationshould lead to transfer between procedurally different shortcutsthat are based on the same mathematical principle and thereforelikely are both associated to the respective conceptual knowledge.For the case of commutativity, we tested whether different strate-gies that are based on the same principle trigger each other via theconcept and so could support flexibility in strategy use. Accordingto the adaptive expertise metaphor (e.g., Hatano, 1988; Star andRittle-Johnson, 2008; Verschaffel et al., 2009) children first of allneed to spontaneously recognize where knowledge can be applied.

Experiment 1 provided first evidence that children who areprovided an opportunity to spontaneously spot and apply oneshortcut might be more inclined to search for and use a sec-ond shortcut, once the first one no longer applies. This isin line with the suggestion to differentiate between (a) quick

and accurate routine-based solving from (b) an adaptive useof solution strategies, which draws upon conceptual under-standing (Hatano, 1988). Experiment 2 verified that transferoccurred from one shortcut to another. It furthermore speci-fied that this transfer effect was not only based on motivation.While we obtained transfer (at least in second graders) fromone commutativity-based shortcut to another commutativity-based shortcut, no transfer was observed between inversionand commutativity. Thus, our results are in line with the viewthat links between different elements of procedural knowl-edge and potentially conceptual knowledge (compare Haideret al., 2014) are used to spontaneously spot and apply shortcutoptions.

Several studies on commutativity have shown that childrenhave at least some understanding of the concept of commutativitybefore entering school (Siegler, 1989; Resnick, 1992; Cowan andRenton, 1996; Wilkins et al., 2001; Canobi et al., 2003) and alreadyfirst graders seem to understand the commutativity principle(Canobi et al., 2002). We thus focused on triggering the usage ofknowledge rather than knowledge acquisition as such. In primaryschool, children should link different strategies based on the sameconcept and develop the ability to select an efficient strategy forthe current problem (Verschaffel et al., 2009). As implied by theseauthors in the adaptive expertise metaphor, the learner should beable to spot and apply options for a shortcut independently with-out having to rely on instruction or explicit cues. In a similar vein,research on skill acquisition and expertise stresses the importanceof linking perceptual skills and principle-knowledge in order tobe able to spontaneously spot and apply shortcuts (e.g., Gentnerand Toupin, 1986; Koedinger and Anderson, 1990; Haider andFrensch, 1996; Anderson and Schunn, 2000; Bilalic et al., 2008a;Frensch and Haider, 2008). Adaptive strategy use can be regardedas the ability to select procedures that can simplify the solution ofa problem (Selter, 2009). In the end the person should be fasterand/or the solution should be more accurate. Strategy use canbe seen as an indicator for the state of development of a mathe-matical concept. Adaptive strategy use necessitates shifts between:(a) calculating problems in the general mode (b) investing sometime and effort to search for shortcut options, and (c) using ashortcut option. We are interested in factors that can tip the bal-ance on the exploitation-exploration continuum. Experts knowwhen to search for a new shortcut strategy and when not, childrenhave to learn how much time and effort they want to spend forsearching. Teachers etc. cannot sustainably take over the regula-tion of this dilemma calculating in standard way or flexible changestrategies—they can only help children to calibrate the balancebetween flexibility vs. stability (or exploration vs. exploitation).






We have to acknowledge that the effects of spontaneously usinga shortcut were small in many cases of the current experimentsand the variability across students was large. This is to be expectedwhen taking into account the difference between competenceand performance (i.e., principle knowledge and application).Larger estimates of both procedural and conceptual knowledgehave been obtained when knowledge was probed more directly(Prather and Alibali, 2009). Direct probing, however, does con-vey to the students that and which shortcut options exist. Itis therefore not suitable when trying to measure the extentto which knowledge about a mathematical principle is appliedspontaneously (cf. Haider et al., 2014). In addition, Robinsonand Dubé (2012) have suggested that personality characteris-tics bridge between knowledge and application. They arguedthat some children have more positive attitudes toward acceptingstrategies that are highly efficient but are novel to their cur-rent strategy repertoire of algorithmic approaches. In a similarvein, (Guerrero and Palomaa, 2012) highlighted that some chil-dren change their strategies during calculation while some donot. Furthermore, children change their strategies for differentreasons. It is not always the goal to choose the most efficientstrategy (Newton et al., 2010) suggested that flexibility involvesthe use of strategies, which are considered the most appropri-ate for a given problem. They also discussed what “appropriate”means. It could be the most efficient or the most understand-able strategy in a given situation. Which strategy in general isused depends on the problem, the numbers presented and othercontextual, developmental, or personal factors (Newton et al.,2010; Guerrero and Palomaa, 2012). An U-shaped relationshipbetween knowledge/understanding and variety of strategy usesuggests that novices as well as experts may use a large vari-ety of strategies (Siegler and Jenkins, 1989; Dowker et al., 1996).Experts like mathematic students used large numbers of appro-priate strategies (Dowker et al., 1996) whereas children (novices)may use a large variety of appropriate and inappropriate strate-gies, because they have not yet acquired a small set of well-learnedstrategies (Dowker et al., 1996). In contrast to this assumptionNewton et al. (2010) argued that low achieving students mightbe particularly appreciative and excited about a focus on multiplestrategies to compare the possible ways to solve the problem andmaximize the accuracy. Although the idea is prominent that aneducational approach for low achieving children should promoteroutine mastery of a single well-thought solution strategy for agiven type of problems (e.g.,Woodward and Baxter, 1997; Baxteret al., 2001). Future work should explore how students at differ-ent ability levels profit from sequences of problems allowing fordifferent shortcuts based on the same mathematical principle.

In order to optimize the chances to measure spontaneous (i.e.,no cues and no instruction) recruitment of knowledge aboutthe commutativity principle we chose a paper-and-pencil test inthe classroom in Experiment 2. Our informal observations sug-gest that children taking part in an eye tracking study on mentalarithmetic appreciate that the measurement is (not only) aboutwhether they solve the problems correctly, but also on how theysolve them. The paper-and-pencil method was closer to usualtest situations in the classroom. Children focused on being fastand accurate rather than on the fact that someone might be

trying to assess how they solved the problems. Verschaffel et al.(2009) highlighted the importance of ecological validity for stud-ies on adaptive expertise. We suggest that trial-by-trial processmeasures (as in our eye tracking experiment) and ecologicallyvalid but less sensitive methods (as in Experiment 2) should becombined to convey the full picture. For instance, eye trackingcan help to figure out whether increased time demands after achange in shortcut option reflect prolonged solution times or,alternatively, a mixture of prolonged solution times plus timeinvested in search for alternative shortcut options. Potentially,learners at different levels of expertise might differ in both theefficiency in spotting shortcuts as well as in using them. Forinstance, third graders might have discovered the options forthe addends-compare shortcut relatively quickly even without afitting warm-up condition.

In line with the research on adaptive expertise Verschaffel et al.(2009) or Star and Rittle-Johnson (2008) defined flexibility inproblem solving as knowledge of multiple strategies and theirrelative efficiency. In addition to weighing different strategiesaccording to their efficiency, students need to weigh the potentialcosts and benefits of flexible strategy usage. There are time costsof switching between strategies, once a shortcut option has beendiscovered (Lemaire and Lecacheur, 2010). Luwel et al. (2009)found longer response times but no reduced accuracy and thesize of these switching costs varied as a function of the associa-tive strength between a strategy and a particular problem. Moreimportantly, there is a dilemma between (a) investing time andattention in order to spot potential shortcut options that might ormight not exist and (b) using processing strategies readily avail-able (e.g., Jepma and Nieuwenhuis, 2011). Thus, process measuresthat provide evidence on when, how and to what extent stu-dents invest in spotting and applying shortcuts (Haider and Rose,2007) are necessary in order to better understand the bases ofthe transfer effect observed in Experiment 2. To illustrate thesearch process, we additionally used eye tracking assessment inthe Experiment 1. On the one hand this is a more specific methodthan paper pencil and on the other hand we could measure theshift of attention. The eye tracking results are in line with the viewof (Robinson and LeFevre, 2012). For discovering new strategies,children need to shift their attention to the relevant part of theproblem. The eye movement patterns were different for the dif-ferent shortcut strategies and fit to the points of interests of theaccording strategies.

ACKNOWLEDGMENTSThis work was supported by Grant FR 1471/12-1 from theDeutsche Forschungsgemeinschaft (DFG) as well as by theBerlin Cluster of Excellence Image Knowledge Gestaltung (www.interdisciplinary-laboratory.hu-berlin.de). Some of the resultswere presented at the Conference of Experimental Psychologists2013 in Vienna, Austria. We thank Christine Arndt, KateKönnecke and Maria Wirth for help with data collection.

SUPPLEMENTARY MATERIALThe Supplementary Material for this article can be foundonline at: http://www.frontiersin.org/journal/10.3389/fpsyg.2014.00556/abstract


https://www.interdisciplinary-laboratory.hu-berlin.de

https://www.interdisciplinary-laboratory.hu-berlin.de






REFERENCESAnderson, J. R., and Schunn, C. D. (2000). “Implications of the ACT-R learning

theory: no magic bullets,” in Advances in Instructional Psychology: EducationalDesign and Cognitive Science, ed R. Glaser (Mahwah, NJ: Erlbaum), 1–33.

Baroody, A. J., and Gannon, K. E. (1984). The development of the commutativ-ity principle and economical addition strategies. Cogn. Instr. 1, 321–339. doi:10.1207/s1532690xci0103_3

Baroody, A. J., Ginsburg, H. P., and Waxman, B. (1983). Children’s use of mathe-matical structure. J. Res. Math. Educ. 14, 156–168. doi: 10.2307/748379

Baxter, J. A., Woodward, J., and Olson, D. (2001). Effects of reform-based mathe-matics instruction on low achievers in five third-grade classrooms. Elem. Sch. J.101, 529–547. doi: 10.1086/499686

Bilalic, M., and McLeod, P. (2014). Why good thoughts block better ones. Sci. Am.310, 74–79. doi: 10.1038/scientificamerican0314-74

Bilalic, M., McLeod, P., and Gobet, F. (2008a). Inflexibility of experts—reality ormyth? Quantifying the Einstellung effect in chess masters. Cogn. Psychol. 56,73–102. doi: 10.1016/j.cogpsych.2007.02.001


Byrnes, J. P., and Wasik, B. A. (1991). Role of conceptual knowledge in math-ematical procedural learning. Dev. Psychol. 27, 777–786. doi: 10.1037/0012-1649.27.5.777

Canobi, K. H. (2009). Concept–procedure interactions in children’s addition andsubtraction. J. Exp. Child Psychol. 102, 131–149. doi: 10.1016/j.jecp.2008.07.008

Canobi, K. H., Reeve, R. A., and Pattison, P. E. (2002). Young chil-dren’s understanding of addition concepts. Educ. Psychol. 22, 513–532. doi:10.1080/0144341022000023608

Canobi, K. H., Reeve, R. A., and Pattison, P. E. (2003). Patterns of knowledge inchildren’s addition. Dev. Psychol. 39, 521–534. doi: 10.1037/0012-1649.39.3.521

Cowan, R., and Renton, M. (1996). Do they know what they are doing? Children’suse of economical addition strategies and knowledge of commutativity. Educ.Psychol. 16, 407–420. doi: 10.1080/0144341960160405

Dowker, A. (2014). Young children’s use of derived fact strategies for addition andsubtraction. Front. Hum. Neurosci. 7:924. doi: 10.3389/fnhum.2013.00924

Dowker, A., Flood, A., Griffiths, H., Harriss, L., and Hook, L. (1996).Estimation strategies of four groups. Math. Cogn. 2, 113–135. doi:10.1080/135467996387499

Dubé, A. K., and Robinson, K. M. (2010). The relationship between adults’ con-ceptual understanding of inversion and associativity. Can. J. Exp. Psychol. 64,60–66. doi: 10.1037/a0017756

ErEl, H., and Meiran, N. (2011). Mindset changes lead to drastic impairments inrule finding. Cognition 119, 149–165. doi: 10.1016/j.cognition.2011.01.002

Ericsson, K. A. (2008). Deliberate practice and acquisition of expert performance:a general overview. Acad. Emerg. Med. 15, 988–994. doi: 10.1111/j.1553-2712.2008.00227.x

Ericsson, K. A., and Lehmann, A. C. (1996). Expert and exceptional performance:evidence of maximal adaptation to task constraints. Annu. Rev. Psychol. 47,273–305. doi: 10.1146/annurev.psych.47.1.273

Frensch, P. A., and Haider, H. (2008). “2.31—transfer and expertise: the search foridentical elements,” in Learning and Memory: A Comprehensive Reference, ed H.B. John (Oxford: Academic Press), 579–596. Available online at: http://www.

sciencedirect.com/science/article/pii/B9780123705099001777Gaschler, R., Vaterrodt, B., Frensch, P. A., Eichler, A., and Haider, H. (2013).

Spontaneous usage of different shortcuts based on the commutativity principle.PLoS ONE 8:e74972. doi: 10.1371/journal.pone.0074972

Gentner, D., and Toupin, C. (1986). Systematicity and surface similarity inthe development of analogy. Cogn. Sci. 10, 277–300. doi: 10.1016/S0364-0213(86)80019-2

Guerrero, S. M., and Palomaa, K. (2012). First-grade methods of single-digitaddition with two or more addends. J. Res. Child. Educ. 26, 1–17. doi:10.1080/02568543.2012.633841

Haider, H., Eichler, A., Hansen, S., Vaterrodt, B., Gaschler, R., and Frensch, P.(2014). How we use what we learn in Math: an integrative account of the devel-opment of commutativity. Frontline Learn. Res. 2, 1–21. doi: 10.14786/flr.v2i1.37

Haider, H., and Frensch, P. A. (1996). The Role of Information Reduction in SkillAcquisition. Cogn. Psychol. 30, 304–337. doi: 10.1006/cogp.1996.0009

Haider, H., and Rose, M. (2007). How to investigate insight: a proposal. Methods42, 49–57. doi: 10.1016/j.ymeth.2006.12.004

Hannula, M. M., and Lehtinen, E. (2005). Spontaneous focusing on numeros-ity and mathematical skills of young children. Learn. Instr. 15, 237–256. doi:10.1016/j.learninstruc.2005.04.005

Hannula, M. M., Lepola, J., and Lehtinen, E. (2010). Spontaneous focusing onnumerosity as a domain-specific predictor of arithmetical skills. J. Exp. ChildPsychol. 107, 394–406. doi: 10.1016/j.jecp.2010.06.004

Harel, A., Kravitz, D., and Baker, C. I. (2013). Beyond perceptual expertise: revis-iting the neural substrates of expert object recognition. Front. Hum. Neurosci.7:885. doi: 10.3389/fnhum.2013.00885

Hatano, G. (1988). Social and motivational bases for mathematical understanding.New Dir. Child Adolesc. Dev. 1988, 55–70. doi: 10.1002/cd.23219884105

Hatano, G., and Oura, Y. (2003). Commentary: reconceptualizing schoollearning using insight from expertise research. Educ. Res. 32, 26–29. doi:10.3102/0013189X032008026

Hiebert, J., and Wearne, D. (1996). Instruction, understanding, and skillin multidigit addition and subtraction. Cogn. Instr. 14, 251–283. doi:10.1207/s1532690xci1403_1

Jepma, M., and Nieuwenhuis, S. (2011). Pupil diameter predicts changes inthe exploration–exploitation trade-off: evidence for the adaptive gain theory.J. Cogn. Neurosci. 23, 1587–1596. doi: 10.1162/jocn.2010.21548

Koedinger, K. R., and Anderson, J. R. (1990). Abstract planning and percep-tual chunks: elements of expertise in geometry. Cogn. Sci. 14, 511–550. doi:10.1207/s15516709cog1404_2

Lemaire, P., and Lecacheur, M. (2010). Strategy switch costs in arithmetic problemsolving. Mem. Cogn. 38, 322–332. doi: 10.3758/MC.38.3.322

Leys, C., Klein, O., Bernard, P., and Licata, L. (2013). Detecting outliers:do not use standard deviation around the mean, use absolute deviationaround the median. J. Exp. Soc. Psychol. 49, 764–766. doi: 10.1016/j.jesp.2013.03.013

Logan, G. D. (1988). Toward an instance theory of automatization. Psychol. Rev. 95,492–527. doi: 10.1037/0033-295X.95.4.492

Luchins, A. S. (1942). “Mechanization in problem solving: the effect of einstellung,”in Psychological Monographs, Vols. 1–6, Vol. 54, ed J. F. Dashiell (Illinois, IL: TheAmerican Psychological Association), 1–95.

Luwel, K., Schillemans, V., Onghena, P., and Verschaffel, L. (2009). Does switch-ing between strategies within the same task involve a cost? Br. J. Psychol. 100,753–771. doi: 10.1348/000712609X402801

McMullen, J. A., Hannula-Sormunen, M. M., and Lehtinen, E. (2011). “Young chil-dren’s spontaneous focusing on quantitative aspects and their verbalizationsof their quantitative reasoning,” in Proceedings of the 35th Conference of theInternational Group for the Psychology of Mathematics Education, Vol. 3 (Ankara:PME), 217–224.

Newton, K. J., Star, J. R., and Lynch, K. (2010). Understanding the development offlexibility in struggling algebra students. Math. Think. Learn. 12, 282–305. doi:10.1080/10986065.2010.482150

Palmeri, T. J., Wong, A. C.-N., and Gauthier, I. (2004). Computational approachesto the development of perceptual expertise. Trends Cogn. Sci. 8, 378–386. doi:10.1016/j.tics.2004.06.001

Prather, R. W., and Alibali, M. W. (2009). The development of arithmetic principleknowledge: how do we know what learners know? Dev. Rev. 29, 221–248. doi:10.1016/j.dr.2009.09.001

Resnick, L. B. (1992). “From protoquantities to operators: building mathematicalcompetence on a foundation of everyday knowledge,” in Analysis of ArithmeticFor Mathematics Teaching, eds G. Leinhardt, R. Putnam, and R. A. Hattrup(Hillsdale, NJ: L. Erlbaum Associates), 1–92.

Rittle-Johnson, B., Siegler, R. S., and Alibali, M. W. (2001). Developing conceptualunderstanding and procedural skill in mathematics: an iterative process. J. Educ.Psychol. 93, 346–362. doi: 10.1037/0022-0663.93.2.346

Robinson, K., and LeFevre, J.-A. (2012). The inverse relation between multiplica-tion and division: concepts, procedures, and a cognitive framework. Educ. Stud.Math. 79, 409–428. doi: 10.1007/s10649-011-9330-5

Robinson, K. M., and Dubé, A. K. (2009). Children’s understanding of addi-tion and subtraction concepts. J. Exp. Child Psychol. 103, 532–545. doi:10.1016/j.jecp.2008.12.002

Robinson, K. M., and Dubé, A. K. (2012). Children’s use of arithmetic short-cuts: the role of attitudes in strategy choice. Child Dev. Res. 2012, 10. doi:10.1155/2012/459385

Rosenberg-Lee, M., Barth, M., and Menon, V. (2011). What difference does a yearof schooling make? Maturation of brain response and connectivity between 2nd


http://www.sciencedirect.com/science/article/pii/B9780123705099001777

http://www.sciencedirect.com/science/article/pii/B9780123705099001777





and 3rd grades during arithmetic problem solving. Neuroimage 57, 796–808.doi: 10.1016/j.neuroimage.2011.05.013

Selter, C. (2009). Creativity, flexibility, adaptivity, and strategy use in mathematics.ZDM 41, 619–625. doi: 10.1007/s11858-009-0203-7

Siegler, R. S. (1989). Hazards of mental chronometry: an example from chil-dren’s subtraction. J. Educ. Psychol. 81, 497–506. doi: 10.1037/0022-0663.81.4.497

Siegler, R. S., and Jenkins, E. (1989). How Children Discover New Strategies, Vol. 14.Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

Star, J. R., and Rittle-Johnson, B. (2008). Flexibility in problem solving: the caseof equation solving. Learn. Instr. 18, 565–579. doi: 10.1016/j.learninstruc.2007.09.018

Verschaffel, L., De Corte, E., Gielen, I., and Struyf, E. (1994). “Clever rear-rangement strategies in children’s mental arithmetic: a confrontationof eye-movement data and verbal protocols,” in Research on Learningand Instruction of Mathematics in Kindergarten and Primary School,ed J. E. H. Van Luit (Doetinchem: Graviant Publishing Company),153–180.

Verschaffel, L., Luwel, K., Torbeyns, J., and Dooren, W. V. (2009). Conceptualizing,investigating, and enhancing adaptive expertise in elementary mathemat-ics education. Eur. J. Psychol. Educ. 24, 335–359. doi: 10.1007/BF03174765

Waldmann, M. (2006). “Konzepte und Kategorien,” in Handbuch der Psychologie,eds J. Funke and P. A. Frensch (Göttingen: Hogrefe Verlag), 283–293.

Wilkins, J. L., Baroody, A. J., and Tiilikainen, S. (2001). Kindergartners’ under-standing of additive commutativity within the context of word problems. J. Exp.Child Psychol. 79, 23–36. doi: 10.1006/jecp.2000.2580

Woodward, J., and Baxter, J. A. (1997). The effects of an innovative approachto mathematics on academically low-achieving students in inclusive settings.Except. Child. 63, 373–388.


Received: 26 March 2014; paper pending published: 16 April 2014; accepted: 19 May2014; published online: 10 June 2014.Citation: Godau C, Haider H, Hansen S, Schubert T, Frensch PA and GaschlerR (2014) Spontaneously spotting and applying shortcuts in arithmetic—a primaryschool perspective on expertise. Front. Psychol. 5:556. doi: 10.3389/fpsyg.2014.00556This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Godau, Haider, Hansen, Schubert, Frensch and Gaschler. Thisis an open-access article distributed under the terms of the Creative CommonsAttribution License (CC BY). The use, distribution or reproduction in other forums ispermitted, provided the original author(s) or licensor are credited and that the originalpublication in this journal is cited, in accordance with accepted academic practice. Nouse, distribution or reproduction is permitted which does not comply with these terms.












ORIGINAL RESEARCH ARTICLEpublished: 02 July 2014

doi: 10.3389/fpsyg.2014.00679

The Einstellung effect in anagram problem solving:evidence from eye movementsJessica J. Ellis and Eyal M. Reingold*

Department of Psychology, University of Toronto Mississauga, Mississauga, ON, Canada

Edited by:

Merim Bilalic, Alpen Adria University,Austria

Reviewed by:

Robert Gaschler, UniversitätKoblenz-Landau, GermanyPeter McLeod, Oxford University, UK

*Correspondence:

Eyal M. Reingold, Department ofPsychology, University of TorontoMississauga, South Building,Room 2037B, 3359 Mississauga RoadNorth, Mississauga, ON L5L 1C6,Canadae-mail: [email protected]

The Einstellung effect is the counterintuitive finding that prior experience or domain-specificknowledge can under some circumstances interfere with problem solving performance.This effect has been demonstrated in several domains of expertise including medicine andchess. In the present study we explored this effect in the context of a simplified anagramproblem solving task. Participants solved anagram problems while their eye movementswere monitored. Each problem consisted of six letters: a central three-letter string whoseletters were part of the solution word, and three additional individual letters. Participantswere informed that one of the individual letters was a distractor letter and were asked to finda five-letter solution word. In order to examine the impact of stimulus familiarity on problemsolving performance and eye movements, the central letter string was presented either asa familiar three-letter word, or the letters were rearranged to form a three-letter nonword.Replicating the classic Einstellung effect, overall performance was better for nonword thanword trials. However, participants’ eye movements revealed a more complex pattern ofboth interference and facilitation as a function of the familiarity of the central letter string.Specifically, word trials resulted in shorter viewing times on the central letter string andlonger viewing times on the individual letters than nonword trials. These findings suggestthat while participants were better able to encode and maintain the meaningful word stimuliin working memory, they found it more challenging to integrate the individual letters intothe central letter string when it was presented as a word.

Keywords: eye movements, problem solving, anagrams, Einstellung effect, insight problems

INTRODUCTIONThe concept that stimulus familiarity and previously acquireddomain knowledge might impair problem solving performancehas been referred to by a variety of interrelated terms includingfunctional fixedness, negative transfer, mental set, and Einstel-lung. Functional fixedness refers to cases where familiarity withhabitual uses of objects blocks other uses from being consid-ered. For example, in the classic “candle-box” insight problemintroduced by Duncker (1945) the presented use of the box as acontainer is hypothesized to interfere with the required consid-eration of the box as a shelf for supporting the candle. Similarly,negative transfer refers to the notion that the retrieval of pre-viously acquired stimulus–response associations can impair theestablishment and maintenance of new stimulus-response associ-ations (e.g., Schultz, 1960; Sweller, 1980; Chrysikou and Weisberg,2005; Landrum, 2005; Osman, 2008). Finally, a problem solvingset or mental set refers to the negative impact of prior exposureto similar problems (either pre-experimental or during the exper-iment), which triggers a familiar but inappropriate solution andprevents alternative solutions from being considered. The Ein-stellung effect (Einstellung is German for attitude), which wasoriginally demonstrated by Luchins’ (1942) seminal series of waterjar experiments, constitutes an excellent illustration of the nega-tive impact of a mental set (for a review see Bilalic and McLeod,2014). In this paradigm, habitual approaches to problem solv-ing are induced through exposure to multiple problems that have

similar solution methods. When a problem is subsequently pre-sented for which the habitual solution method is not appropriate,many participants claim that the problem is unsolvable. However,naive participants can find the solution quickly, thus showingthat the problem is not intrinsically difficult and that the diffi-culty experienced by solvers reflects the negative impact of priorexperience.

When considered in the context of human expertise, the ideathat prior experience and stimulus familiarity might interferewith problem solving performance seems at first blush to berather counterintuitive. This is because there is a large body ofresearch demonstrating that stimulus familiarity and domain-specific knowledge acquired through extensive and deliberatepractice underlie the superior performance of experts relative totheir less skilled counterparts (for a review see Ericsson and Char-ness, 1994). However, experts are not immune to the negativeimpact of prior experience and stimulus familiarity as demon-strated in studies of expertise in medicine (de Graaff, 1989;Croskerry, 2003; Gordon and Franklin, 2003) and chess (Saar-iluoma, 1992; Reingold et al., 2001b; Bilalic et al., 2008a,b, 2010;Sheridan and Reingold, 2013; Bilalic and McLeod, 2014). Forexample, Bilalic et al. (2008a) employed eye movement moni-toring to study the Einstellung effect in chess experts. Playerswere required to find a checkmate with the fewest number ofmoves. There were two possible solutions: a familiar five-movesequence and a less well-known three-move sequence (the optimal

www.frontiersin.org July 2014 | Volume 5 | Article 679 | 35












Ellis and Reingold Einstellung effect in anagram problem solving

solution). After identifying the familiar solution, expert chess play-ers reported that they were searching for the optimal one. However,the eye movement record revealed that their attention continuedto be directed more often towards chess board regions involved inthe familiar rather than the optimal solution. Thus, it appears thatthe Einstellung effect demonstrated in chess experts was due tothe familiar scenario activating a schema in memory that directsattention towards information relevant to itself, and away fromother information (Bilalic et al., 2008a, 2010; Bilalic and McLeod,2014).

In the present study, in order to further investigate the neg-ative influence of stimulus familiarity on problem solving, wemonitored participants’ eye movements while they performed amodified anagram problem solving task that was introduced byEllis et al. (2011). Anagram tasks provide a unique opportunity tostudy the Einstellung effect in a domain of expertise possessedby most adults, that is, their familiarity with words. In addi-tion, unlike most problem solving tasks that were employed tostudy the Einstellung effect, the use of anagrams allows for thecreation of a large number of independent trials in which anEinstellung effect might occur. Anagrams have long been usedto study insight problem solving (for a review see Ellis et al.,2011) as well as to demonstrate the negative impact of a men-tal set on problem solving performance (e.g., Rees and Israel,1935; Maltzman and Morrisett, 1952; Kaplan and Schoenfeld,1966; Juola and Hergenhahn, 1967). In particular, it has beenestablished (e.g., Beilin and Horn, 1962; Ekstrand and Domi-nowski, 1965, 1968; Tresselt and Mayzner, 1965; Mayzner andTresselt, 1966) that solution rates are lower and response timesare slower when the solution word (e.g., HEART) is scram-bled to create a word anagram (e.g., EARTH) than a nonwordanagram (e.g., THREA). It is likely that the familiar word ana-gram produces activation (orthographic, phonological, lexical,and/or semantic) that is irrelevant to the solution and hin-ders the decomposition and restructuring operations that arerequired to produce the solution word (e.g., Hollingworth, 1935;Devnich, 1937).

The present investigation involved eye movement monitor-ing during anagram problem solving. As illustrated in Figure 1,the anagram task we used consisted of six uppercase letters: acentrally located three-letter string, plus three individual letterspositioned above and to either side of the central letter string.The central letter string could be arranged either as a familiarthree-letter word, or as a meaningless string of three letters. Par-ticipants were asked to produce a five-letter solution word, andwere informed that one of the individual letters was a distractorletter that was not part of the solution word. Using a similar ana-gram task, Ellis et al. (2011) reported that near the beginning oftrials, viewing times on the distractor and solution letters wereindistinguishable, meaning that participants did not immediatelyperceive that the distractor letter was irrelevant to the solution.Towards the end of trials, viewing times on the distractor letterdecreased relative to the solution letters, indicating that partialsolution knowledge had developed, and this change occurred sev-eral seconds prior to solution. Importantly, the pattern of viewingtimes was the same regardless of whether or not participantsreported a subjective experience of insight upon solution, thereby

FIGURE 1 | Examples of the anagram problem stimulus display, shown

with the central letter string arranged as a word (A) and as a

meaningless letter string (B), and an illustration of the circular interest

areas within which fixations were assigned to the individual letters or

the central letter string, shown here as Xs (C). The solution to theanagram problems is TOUCH.

demonstrating a dissociation between the subjective experience ofinsight and the objective accumulation of solution knowledge. Inthe present study we expected to replicate the pattern reportedby Ellis et al. (2011). In addition, based on previous findingswith anagrams (e.g., Beilin and Horn, 1962; Ekstrand and Domi-nowski, 1965, 1968; Tresselt and Mayzner, 1965; Mayzner and

Frontiers in Psychology | Cognition July 2014 | Volume 5 | Article 679 | 36




Tresselt, 1966) we predicted better problem solving performancewhen the central letter string was presented as a nonword thana word. Finally, we explored differences in the pattern of look-ing behavior as a function of the familiarity of the central letterstring.

MATERIALS AND METHODSPARTICIPANTSSixty undergraduates from the University of Toronto Mississaugaparticipated in exchange for partial course credit or $10. All partic-ipants had normal or corrected-to-normal vision and were fluentEnglish speakers.

APPARATUSAn SR Research EyeLink 1000 eye tracking system was usedto record participants’ eye movements with a sampling rate of1000 Hz. The stimuli were displayed on a 19-inch Viewsonicmonitor with a refresh rate of 75 Hz and a screen resolution of1024 × 768. Participants were seated 60 cm from the display andused a chinrest with a head support to minimize head movement.Following calibration, gaze-position error was less than 0.5◦.

MATERIALSAnagram problems consisted of six uppercase letters: a cen-trally located three-letter string, plus three individual letterspositioned above and to either side of the central letter string(see Figure 1 for examples). All three letters in the central let-ter string belonged to the solution word, while only two of theindividual letters belonged to the solution word, with the thirdbeing a randomly placed distractor letter. The task was to com-bine two of the three individual letters with the central letterstring to create a five-letter solution word. Each anagram prob-lem had only one possible solution, meaning that the distractorletter did not allow for the formation of any alternative five-letterwords.

Each anagram problem could be presented in either “word” or“nonword” condition. In the “word” condition, the central letterstring consisted of a three-letter word, while in the “nonword”condition, the central letter string consisted of a scrambled non-word version of those same three letters (see Figure 1). For eachanagram problem, the identity and location of the three individ-ual letters was the same across both conditions, such that the onlydifference between a given anagram problem in the two condi-tions was the configuration of the central letter string as a word ornonword. In the “word” condition, the central letter string wordshad a mean frequency of 435 per million (SD = 1243) accordingto Brysbaert and New (2009).

Solution words were made up of five unique letters, alwaysbegan with a consonant, and contained either one vowel (33%of problems) or two vowels (67% of problems). Solution wordshad a mean frequency of 175 per million (SD = 396) accordingto Brysbaert and New (2009). The central letter string always con-sisted of two consonants and a vowel, as did the three individualletters. In order to remind participants that the three letters inthe central letter string must always be included in the solutionword, these letters were displayed in green, in a slightly smallerand bolder font than the three individual letters, which were

displayed in black. Each anagram problem subtended approx-imately five visual degrees in height and 14 visual degrees inwidth.

The location of the individual distractor letter was counter-balanced across anagram problems. In an attempt to avoid anya priori bias away from the distractor letter, we matched thedistractor letter with the other two individual solution lettersin terms of letter frequency (averaged across all five possi-ble letter positions within the solution word) using tables byMayzner and Tresselt (1965). Across all experimental anagrams,the mean frequency of the distractor letter was no differentfrom the mean frequency of the individual solution letters (dis-tractor M = 193, SD = 91, solution M = 199, SD = 115,t < 1).

PROCEDUREParticipants completed six practice trials followed by 72 experi-mental trials. Half the anagram problems were presented in the“word” condition and half were presented in the “nonword” con-dition, and both anagram order and central letter string type wererandomized for each participant. Across participants, each ana-gram problem was presented an equal number of times in the“word” and “nonword” conditions.

Every trial began on a blank screen with a central fixa-tion cross. After 1000 ms, the anagram problem appeared andremained on the screen until a response was made, or untilthe trial timed out after 45 s. Participants were instructedthat speed of responding was of utmost importance and werediscouraged from verifying their solution prior to response,even if that might elicit the occasional incorrect solution.Participants pressed a button on the response pad in orderto respond. The stimulus display then disappeared and par-ticipants verbalized their answer to the experimenter, whoprovided feedback as to whether or not their response wascorrect.

After every trial, participants were asked to classify their sub-jective experience of solving the anagram problem. Participantsselected one of the following options (from Novick and Sherman,2003) by pressing a corresponding button on the response pad.

1 “The solution came to mind suddenly, seemingly out ofnowhere. I have no awareness of having done anything to tryto get the answer.”

2 “I tried various letter arrangements in order to solve the ana-gram, but none of them seemed to work. Then the solutioncame to mind suddenly.”

3 “I tried various letter arrangements in order to solve the ana-gram. I was able to build on one of these arrangements to workout the solution step by step.”

4 “I did not solve the anagram.”

We considered options 1 and 2 to describe subjective experi-ences of insight, and labeled all trials where participants selectedoption 1 or 2 as “popout” trials. Option 3 does not describea subjective experience of insight, so trials where participantsselected option 3 were labeled “non-popout” trials. Participantsmade another button press to advance to the next trial at theirown pace.





RESULTSOur main focus in this experiment involved examining the effectof central letter string type (word vs. nonword) on mean task per-formance and eye movement measures. However, we also wantedto ensure that manipulating central letter string type did not alterthe nature of the problem solving task as compared to prior find-ings (Ellis et al., 2011). Accordingly, while we primarily focus onthe differences between word and nonword trials, we also exam-ined any interaction between the familiarity of the central letterstring and participants’ subjective experience of insight (i.e., tri-als in which the solution was experienced as emerging suddenlywere classified as popout trials, whereas trials in which the solu-tion was experienced as gradual were classified as non-popouttrials; see Method section for details). Across participants, 52.7%of trials were correct, 4.2% of trials were incorrect, and 43.1%of trials timed out with no solution. Mean response time forcorrect trials was 14.3 s (SD = 3.3 s), with 69.0% of correct tri-als classified by participants as popout, and 31.0% classified asnon-popout. The effect of central letter string type on overallproblem solving performance is summarized in Table 1. Impor-tantly, response times were significantly slower for word trials thanfor nonword trials. In addition, there was a numerical trend towardlower accuracy for word trials than nonword trials, althoughthis difference did not reach significance. Finally, there was noimpact of stimulus familiarity on the subjective experience ofinsight problem solving, as shown by the virtually identical pro-portion of word trials and nonword trials that were classified aspopout.

Eye movement analyses were performed only on correct trials,and only included fixations that could be assigned to a particularitem in the stimulus display. Specifically, a fixation was assignedto an individual letter or to the central letter string if it fell withina 192 pixel diameter circle around that item (these fixation areasdid not overlap; for an illustration, see Figure 1C). Within cor-rect trials across participants, 69.9% of fixations were assignedto the central letter string, 28.3% of fixations were assigned toone of the three individual letters, and 1.8% of fixations couldnot be assigned to either the central letter string or the individualletters. Assigned fixations were then converted to dwells, wherea dwell is defined as one or more consecutive fixations withinthe same area prior to an eye movement to another area. Asshown in Table 1, corresponding to the slower response timesfor word than nonword trials, there was a greater number ofdwells per trial on both the central letter string and the individualletters for word trials as compared to nonword trials, revealing

the classic negative influence of familiarity on problem solvingperformance.

However, several fine-grained eye movement measures, shownin Figure 2, revealed a more complex pattern of the effects ofcentral letter string type. More specifically, we calculated overallmeans for the following eye movement measures: (a) duration ofthe initial latency on the central letter string (i.e., the interval fromstimulus onset until the first eye movement that exited the centralletter string area); (b) dwell duration on the central letter stringduring subsequent revisits; and (c) dwell duration on the indi-vidual letters. For each eye movement measure, we carried out a2 × 2 ANOVA with subjective report (popout vs. non-popout) andcentral letter string type (word vs. nonword) as independent vari-ables. As can be seen in Figure 2, some eye movement measuresrevealed facilitation for word trials relative to nonword trials, whileother eye movement measures revealed interference. In addition,there were no significant interactions between central letter stringtype and subjective report for any eye movement measure (allFs < 1.45, n.s.), indicating that the differences between word andnonword trials were the same for both popout and non-popouttrials. Specifically, the initial latency on the central letter string wasmuch shorter for word trials than for nonword trials (Figure 2A;F(1,59) = 29.78, p < 0.001), and this processing advantage forwords over nonwords was also present for subsequent dwells onthe central letter string (Figure 2B; F(1,59) = 7.78, p < 0.01).Thisprocessing advantage for word trials is likely due to a workingmemory advantage in encoding and maintaining the central letterstring when it is arranged as a word as compared to a nonword.In marked contrast, dwell duration on individual letters revealeda processing disadvantage for word trials. Specifically, dwells onindividual letters were longer for word trials than for nonwordtrials [Figure 2C; F(1,59) = 25.31, p < 0.001]. This processingdisadvantage might be due to difficulty in integrating the individ-ual letters into the central letter string when it is in the form of aunitary gestalt.

In addition, we contrasted viewing times on the distractor andsolution letters during the first half and second half of trials as afunction of both the familiarity of the central letter string (wordvs. nonword) and the reported subjective experience (popout vs.non-popout). Based on the findings reported by Ellis et al. (2011),viewing times for the distractor letter and the solution letters wereexpected to be the same at the beginning of trials, whereas solu-tion knowledge towards the end of trials should result in lowerviewing times on the distractor letter as compared to the solutionletters. To examine this prediction, we compared the proportion

Table 1 | Mean problem solving and eye movement measures shown overall for correct trials, and separately for word and nonword trials.

Variable Overall Word Nonword Significance

Accuracy (% correct) 52.7 (2.4) 51.7 (2.4) 53.8 (2.6) t (59) = 1.52, n.s.

Response time (s) 14.3 (0.4) 15.3 (0.6) 13.4 (0.4) t (59) = 3.25, p < 0.01

Number of dwells, central 7.39 (0.37) 8.11 (0.41) 6.76 (0.37) t (59) = 4.69, p < 0.001

Number of dwells, letters 9.40 (0.55) 10.45 (0.63) 8.53 (0.53) t (59) = 4.80, p < 0.001

Percentage of trials classified as popout 69.0 (3.4) 68.9 (3.5) 69.2 (3.6) t < 1, n.s.

SEs are shown in parentheses.





FIGURE 2 | Mean eye movement measures for word and nonword

conditions within popout and non-popout trials: initial latency on the

central letter string (A); subsequent dwell duration on the central

letter string (B); and dwell duration on individual letters (C).

of time spent on the distractor letter and the mean of the twosolution letters separately for the first and second half of trials.Accordingly, we conducted 2 × 2 × 2 ANOVAs on the proportionof viewing time with letter type (distractor vs. solution), sub-jective report (popout vs. non-popout) and central letter stringtype (word vs. nonword) as independent variables. As can be seenin Figures 3A–D, for all conditions, there was no difference inthe first half of trials between viewing times on the distractorletter and the solution letters (all ts < 1.50, n.s.). Likewise, theANOVA revealed no significant main effects or interactions (allFs < 2.74, n.s.). In contrast, in the second half of trials for allconditions, a significantly greater proportion of viewing time wasspent on the solution letters as compared to the distractor letter(all ts > 3.83, all ps < 0.001). In this case, the ANOVA revealed asignificant main effect of letter type [F(1,59) = 75.65, p < 0.001]

but no other main effects or interactions approached significance(all Fs < 0.23, n.s.). The lack of any main effect or interactioninvolving central letter string type suggests that the accumulationof solution knowledge prior to insight is independent of the effectsof the familiarity manipulation. Finally, we also examined the pat-tern of looking behavior during the first and second half of trialsin which participants failed to provide a solution for the anagram.As shown in Figures 3E,F, failure to solve anagrams was reflectedby a small but significant tendency for greater viewing time onthe distractor letter relative to the solution letters [F(1,59) = 5.51,p < 0.05]. No other main effect or interaction was significant(all Fs < 2.48, n.s.).

DISCUSSIONThe main goal of the present study was to explore the negativeinfluence of familiarity on performance in a simplified anagramproblem solving task. We replicated prior findings from the ana-gram literature that showed that task performance is poorerwhen anagrams are presented in word form than when theyare presented as scrambled letters (e.g., Beilin and Horn, 1962;Ekstrand and Dominowski, 1965, 1968; Tresselt and Mayzner,1965; Mayzner and Tresselt, 1966). This effect is thought tobe due to the difficulty in breaking the gestalt of the exist-ing word in order to rearrange the letter order and form anew word. However, participants’ eye movements in the presentstudy revealed a more intricate pattern of the effects of stimulusfamiliarity on anagram problem solving, including both interfer-ence and facilitation. Specifically, the present study documentedshorter viewing times on the central letter string when it waspresented in word form than in nonword form, suggesting thatparticipants were better able to encode and maintain the cen-tral letter string in working memory when it was a meaningfulword than when it was a meaningless string of letters. This find-ing is consistent with the well established perceptual encodingand working memory advantage for familiar stimuli relative tounfamiliar stimuli (e.g., Chase and Simon, 1973; Reingold et al.,2001a).

Of more interest is what the eye movement record revealedabout how familiarity interferes with anagram problem solving.Our paradigm demonstrated that the interference of stimulusfamiliarity was due to longer viewing times on the individual let-ters when the central letter string was presented as a word thanwhen it was presented as a nonword. As originally proposed byGestalt psychologists, these longer viewing times on the individ-ual letters might suggest that participants find it more challengingto integrate the individual letters into the central letter stringwhen it is a holistic entity. Supporting evidence for this possibil-ity comes from simple letter-insertion tasks that were introducedby Reingold (1995). For example, similar to the present task, in aletter-insertion task participants were presented with a letter stringthat was either a word (e.g., CASH) or a nonword (e.g., CRAH)and were required to insert one of two alternative letters into theletter string to create a word (e.g., CRASH). Performance wassubstantially better for nonword than word letter strings (Rein-gold, 1995). Thus, consistent with the conceptualization of Gestaltpsychology, it seems intuitive that restructuring and integratingnew elements into a pre-existing holistic representation would be





FIGURE 3 | Proportion of viewing time on the solution letters and the distractor letter in the first and second half of trials for each combination of

central letter string type (word in A,C,E; nonword in B,D,F) and trial type (popout in A,B; non-popout in C,D; unsolved in E,F).

more difficult than integrating items into an unrelated collectionof problem elements. However, more work is required in orderto specify the mechanisms underlying this theoretical possibil-ity, and to explain how familiarity interferes with the integrationprocess.

Finally, a previous eye movement study of the Einstellung effectin chess experts suggested that the activation of familiar schemas inmemory creates perceptual biases towards information that con-firms these schemas and away from information that is requiredto find a less familiar but more optimal solution (e.g., Bilalicet al., 2008a). Our eye movement analysis did not reveal a per-ceptual bias towards the central letter string when it was presentedin word form. This was likely due to the encoding and process-ing advantages that are associated with the familiar central word.These advantages might have allowed problem solvers to eas-ily maintain familiar stimulus information in working memory

while directing their visual attention elsewhere in the stimulusdisplay. However, participants’ processing resources might havebeen captured by the familiar central letter string in ways thatour eye movement methodology could not reveal. Specifically, itmight be that the familiarity of the central letter string causesan unhelpful bias in the search for solution words based on theirrelevant orthographic, phonological, lexical, and/or semanticactivation associated with the central word. Unlike previous find-ings which demonstrated that an exhaustive encoding mental setcould be replaced by a selective encoding strategy which ignoresirrelevant aspects of the stimulus regardless of stimulus familiar-ity (e.g., Gaschler and Frensch, 2007, 2009; see also Dreisbachand Haider, 2008, 2009), in the present study, participants wereseemingly unable to ignore the irrelevant but familiar centralword. This is likely due to the fact that the irrelevant activationcaused by the central word was involuntary in nature and required





processing resources in order to oppose it. Taken together, thepresent findings and prior results indicate that while stimulusfamiliarity and domain knowledge are clearly fundamental toestablishing expertise, these aspects of skilled performance are notwithout their pitfalls when a problem solving scenario requiresflexibility.

ACKNOWLEDGMENTThis research was supported by an NSERC grant to Eyal Reingold.

REFERENCESBeilin, H., and Horn, R. (1962). Transition probability effects in ana-

gram problem solving. J. Exp. Psychol. 63, 514–518. doi: 10.1037/h0042337


Bilalic, M., McLeod, P., and Gobet, F. (2008a). Why good thoughts block better ones:the mechanism of the pernicious Einstellung (set) effect. Cognition 108, 652–661.doi: 10.1016/j.cognition.2008.05.005

Bilalic, M., McLeod, P., and Gobet, F. (2008b). Inflexibility of experts–reality ormyth? Quantifying the Einstellung effect in chess masters. Cogn. Psychol. 56,73–102. doi: 10.1016/j.cogpsych.2007.02.001

Bilalic, M., McLeod, P., and Gobet, F. (2010). The mechanism of the Einstellung (set)effect: a pervasive source of cognitive bias. Curr. Dir. Psychol. Sci. 19, 111–115.doi: 10.1177/0963721410363571

Brysbaert, M., and New, B. (2009). Moving beyond Kucera and Francis: a criticalevaluation of current word frequency norms and the introduction of a new andimproved word frequency measure for American English. Behav. Res. Methods 41,977–990. doi: 10.3758/BRM.41.4.977

Chase, W. G., and Simon, H. A. (1973). Perception in Chess. Cogn. Psychol. 4, 55–81.doi: 10.1016/0010-0285(73)90004-2

Chrysikou, E. G., and Weisberg, R. W. (2005). Following the wrong foot-steps: fixation effects of pictorial examples in a design problem-solving task.J. Exp. Psychol. Learn. Mem. Cogn. 31, 1134–1148. doi: 10.1037/0278-7393.31.5.1134

Croskerry, P. (2003). The importance of cognitive errors in diagnosis and strategiesto minimize them. Acad. Med. 78, 775–780. doi: 10.1097/00001888-200308000-00003

de Graaff, E. (1989). A test of medical problem-solving scored by nurses anddoctors: the handicap of expertise. Med. Educ. 23, 381–386. doi: 10.1111/j.1365-2923.1989.tb01564.x

Devnich, G. E. (1937). Words as ‘Gestalten.’ J. Exp. Psychol. 20, 297–300. doi:10.1037/h0060468

Dreisbach, G., and Haider, H. (2008). That’s what task sets are for: shielding againstirrelevant information. Psychol. Res. 72, 355–361. doi: 10.1007/s00426-007-0131-5

Dreisbach, G., and Haider, H. (2009). How task representations guide attention:further evidence for the shielding function of task sets. J. Exp. Psychol. Learn.Mem. Cogn. 35, 477–486. doi: 10.1037/a0014647

Duncker, K. (1945). On problem-solving. Psychol. Monogr. 58, i–113. doi:10.1037/h0093599

Ekstrand, B. R., and Dominowski, R. L. (1965). Solving words as anagrams. Psychon.Sci. 2, 239–240. doi: 10.3758/BF03343425

Ekstrand, B. R., and Dominowski, R. L. (1968). Solving words as ana-grams: II. A clarification. J. Exp. Psychol. 77, 552–558. doi: 10.1037/h0026073

Ellis, J. J., Glaholt, M. G., and Reingold, E. M. (2011). Eye movements revealsolution knowledge prior to insight. Conscious. Cogn. 20, 768–776. doi:10.1016/j.concog.2010.12.007

Ericsson, K. A., and Charness, N. (1994). Expert performance: its structure andacquisition. Am. Psychol. 49, 725–747. doi: 10.1037/0003-066X.49.8.725

Gaschler, R., and Frensch, P. A. (2007). Is information reduction an item-specific or an item-general process? Int. J. Psychol. 42, 218–228. doi:10.1080/00207590701396526

Gaschler, R., and Frensch, P. A. (2009). When vaccinating against informationreduction works and when it does not work. Psychol. Stud. 54, 43–53. doi:10.1007/s12646-009-0006-5

Gordon, R., and Franklin, N. (2003). Cognitive underpinnings of diag-nostic error. Acad. Med. 78, 782. doi: 10.1097/00001888-200308000-00005

Hollingworth, H. L. (1935). The conditions of verbal configuration. J. Exp. Psychol.18, 299–306. doi: 10.1037/h0057110

Juola, J. F., and Hergenhahn, B. R. (1967). Effects of overtraining on the estab-lishment of mental set in anagram solving. Psychon. Sci. 9, 539–540. doi:10.3758/BF03327878

Kaplan, I. T., and Schoenfeld, W. N. (1966). Oculomotor patterns during thesolution of visually displayed anagrams. J. Exp. Psychol. 72, 447–451. doi:10.1037/h0023632

Landrum, R. E. (2005). Production of negative transfer in a problem-solving task.Psychol. Rep. 97, 861–866.

Luchins, A. S. (1942). Mechanization in problem solving: the effect of Einstellung.Psychol. Monogr. 54, i–95. doi: 10.1037/h0093502

Maltzman, I., and Morrisett, L. (1952). Different strengths of set inthe solution of anagrams. J. Exp. Psychol. 44, 242–246. doi: 10.1037/h0056022

Mayzner, M. S., and Tresselt, M. E. (1965). Tables of single-letter and digramfrequency counts for various word-length and letter-position combinations.Psychon. Monogr. Suppl. 1, 13–32.

Mayzner, M. S., and Tresselt, M. E. (1966). Anagram solution times: a func-tion of multiple-solution anagrams. J. Exp. Psychol. 71, 66–73. doi: 10.1037/h0022665

Novick, L. R., and Sherman, S. J. (2003). On the nature of insight solutions: evidencefrom skill differences in anagram solution. Q. J. Exp. Psychol. 56A, 351–382. doi:10.1080/02724980244000288

Osman, M. (2008). Positive transfer and negative transfer/antilearning of problem-solving skills. J. Exp. Psychol. Gen. 137, 97–115. doi: 10.1037/0096-3445.137.1.97

Rees, H. J., and Israel, H. E. (1935). An investigation of the establishment andoperation of mental sets. Psychol. Monogr. 6, 1–26. doi: 10.1037/h0093375

Reingold, E. M. (1995). Facilitation and interference in indirect/implicit memorytests and in the process dissociation paradigm: the letter insertion and the letterdeletion tasks. Conscious. Cogn. 4, 459–482. doi: 10.1006/ccog.1995.1051

Reingold, E. M., Charness, N., Pomplun, M., and Stampe, D. M. (2001a). Visual spanin expert chess players: evidence from eye movements. Psychol. Sci. 12, 48–55.doi: 10.1111/1467-9280.00309

Reingold, E. M., Charness, N., Schultetus, R. S., and Stampe, D. M.(2001b). Perceptual automaticity in expert chess players: parallel encod-ing of chess relations. Psychon. Bull. Rev. 8, 504–510. doi: 10.3758/BF03196185

Saariluoma, P. (1992). Error in chess: the apperception-restructuring view. Psychol.Res. 54, 17–26. doi: 10.1007/BF01359219

Schultz, R. W. (1960). Problem solving behavior and transfer. Harv. Educ. Rev. 30,61–77.

Sheridan, H., and Reingold, E. M. (2013). The mechanisms and boundary condi-tions of the Einstellung effect in chess: evidence from eye movements. PLoS ONE8:e75796. doi: 10.1371/journal.pone.0075796

Sweller, J. (1980). Transfer effects in a problem solving context. Q. J. Exp. Psychol.32, 233–239. doi: 10.1080/14640748008401159

Tresselt, M. E., and Mayzner, M. S. (1965). Anagram solution times: a function ofindividual differences in stored digram frequencies. J. Exp. Psychol. 70, 606–610.doi: 10.1037/h0022667


Received: 27 April 2014; accepted: 12 June 2014; published online: 02 July 2014.Citation: Ellis JJ and Reingold EM (2014) The Einstellung effect in anagram problemsolving: evidence from eye movements. Front. Psychol. 5:679. doi: 10.3389/fpsyg.2014.00679This article was submitted to Cognition, a section of the journal Frontiers in Psychology.Copyright © 2014 Ellis and Reingold. This is an open-access article distributed underthe terms of the Creative Commons Attribution License (CC BY). The use, distributionor reproduction in other forums is permitted, provided the original author(s) or licensorare credited and that the original publication in this journal is cited, in accordance withaccepted academic practice. No use, distribution or reproduction is permitted whichdoes not comply with these terms.







ORIGINAL RESEARCH ARTICLEpublished: 10 September 2014doi: 10.3389/fpsyg.2014.00955

Interference between face and non-face domains ofperceptual expertise: a replication and extensionKim M. Curby 1* and Isabel Gauthier 2

1 Department of Psychology, Macquarie University, Sydney, NSW, Australia2 Department of Psychology, Vanderbilt University, Nashville, TN, USA

Edited by:


Reviewed by:

Amy L. Boggan, Young Harris College,USAEva M. Dundas, Carnegie MellonUniversity, USA

*Correspondence:

Kim M. Curby, Department ofPsychology, Macquarie University,Building C3A, Room 409, Sydney,NSW, Australiae-mail: [email protected]

As car expertise increases, so does interference between the visual processing of faces andthat of cars; this suggests performance trade-offs across domains of real-world expertise.Such interference between expert domains has been previously revealed in a relativelycomplex design, interleaving 2-back part-judgment task with faces and cars (Gauthier et al.,2003). However, the basis of this interference is unclear. Experiment 1A replicated thefinding of interference between faces and cars, as a function of car expertise. Experiments1B and 2 investigated the mechanisms underlying this effect by (1) providing baselinemeasures of performance and (2) assessing the specificity of this interference effect.Our findings support the presence of expertise-dependent interference between face andnon-face domains of expertise. However, surprisingly, it is in the condition where facesare processed among cars with a disrupted configuration where expertise has a greaterinfluence on faces. This finding highlights how expertise-related processing changes alsooccur for transformed objects of expertise and that such changes can also drive interferenceacross domains of expertise.

Keywords: perceptual expertise, interference, faces, objects, holistic processing

INTRODUCTIONFace perception is often described as a domain of perceptual exper-tise. Our skill with faces manifests itself across many differenttasks and is often particularly impressive for familiar faces. Forexample, normal adults can recognize familiar faces with accu-racy >90% despite not having seen some of these faces for over35 years (Bahrick et al., 1975). Most people are as fast to categorizean image as a “face” as they are to categorize it at an individuallevel (“Bill Clinton’s face”; Tanaka, 2001). In contrast, observersare much slower to categorize an image of a bird at a similar subor-dinate level—for example, categorizing an animal as a “cardinal”is slower than categorizing it at the basic level, “bird” (Tanakaand Taylor, 1991). But even the processing of unfamiliar facesoutshines our performance with other objects in several respects.For instance, observers can retain more faces in visual short-termmemory than they can other objects (Curby and Gauthier, 2007;Curby et al., 2009). Face processing is also more sensitive to subtlechanges in the spatial-relations between features than object pro-cessing (Haig, 1984; Hosie et al., 1988; Kemp et al., 1990; Bruceet al., 1991).

Our skill with faces is believed to result, at least in part, fromour extensive experience with them (but see Kanwisher, 2000), andto be mediated by the acquisition of a holistic processing strategy(Diamond and Carey, 1977; Richler et al., 2011a). What holisticmeans varies to some extent, as it is sometimes described as asensitivity to configuration, a global (as opposed to local) infor-mation sampling strategy, a situation where perceiving the wholeis greater than the sum of its parts, or integrality of processingfor different dimensions (see Richler et al., 2012 for a review).These different meanings motivate authors to use a number oftasks to compare face and object recognition. One commonly

used meaning of holistic processing is as a failure of selectiveattention. For instance, Young et al. (1987) asked people to namethe identity of part of a face composite, and found they wereunable to do so while ignoring other parts of the composite.When the composite is inverted or the composite parts are mis-aligned, people can more easily selectively attend to a face part.This has been replicated in variations on this original paradigm,such as in matching tasks with unfamiliar faces (Farah et al., 1998;Richler et al., 2008; Curby et al., 2013). Like several other hall-marks of face perception [e.g., the inversion effect; Rossion et al.,2002; Curby et al., 2009; the sensitivity to spatial frequency con-tent; McGugin and Gauthier, 2010; recruitment of the fusiformface area (FFA); Gauthier et al., 2000, 2005; McGugin et al., 2012a],this kind of holistic processing has also been obtained for non-facecategories when expert observers are tested, for instance for carsin car experts (Gauthier et al., 2003; Bukach et al., 2010), chessdisplays in chess experts (Boggan et al., 2012), and novel objectsafter expertise training in the lab (Gauthier and Tarr, 2002; Wonget al., 2009).

Based on the idea that face processing can be understood asa kind of expertise, it has been suggested that it may share moreresources with the processing of other objects of expertise thanwith typical object perception. The logic is simple: if the per-ceptual strategies and neural substrates were found to be similar,this may lead to interference when two categories of expertise areprocessed simultaneously. We originally tested this prediction inan electrophysiological study using the composite paradigm tomeasure holistic processing and a neural marker of expertise, theN170. We recruited participants with a range of car-recognitionskills, from none to extensive, and developed a paradigm in whichparticipants processed faces and cars concurrently (Gauthier et al.,

www.frontiersin.org September 2014 | Volume 5 | Article 955 | 42












Curby and Gauthier Interference across domains of expertise

2003). Participants matched the bottom parts of face and carcomposites, while faces and cars alternated. In this 2-back part-matching task, we were able to measure holistic processing of faceswhen presented in two different contexts: (1) Among normal cars,which car experts were found to process more holistically than carnovices, and (2) among cars with inverted tops, which car expertsdid not process holistically. Therefore, we expected that holisticprocessing of normal cars would compete with that of faces, onlyin car experts. Indeed, we found that faces in the context of nor-mally configured cars were processed less holistically [i.e., therewas less influence from the to-be-ignored (top) part on bottomjudgments] than those presented in the context of cars in a trans-formed configuration (tops inverted). These results suggested afunctional overlap between face and car processing that is relatedto an individual’s level of expertise with cars.

Since our original study, there have been other studies provid-ing evidence of interference between faces and objects of expertiseusing event-related potential (ERP; Rossion et al., 2004, 2007),functional magnetic resonance imaging (fMRI; McGugin et al.,2014), or other behavioral paradigms (McKeeff et al., 2010; McGu-gin et al., 2011). There are also other studies suggesting that theprocessing of faces and words may compete during developmentand influence their lateralization in the brain (Dehaene and Cohen,2011; Dundas et al., 2013). However, Gauthier et al. (2003) is theonly study that looked at functional overlap specifically in holis-tic processing. The goal of the present study was to replicate thisfinding (Simons, 2014) and to explore its underlying mechanismsusing baseline conditions that were not used before.

The measurement of holistic processing is relatively complex,with holistic processing in the composite task quantified using adifference score between two indices of discriminability (each a d′measure that depends on a hit-rate and a false-alarm rate). Thedesign used by Gauthier et al. (2003) not only requires holisticprocessing to be measured for both faces and cars concurrently,but also the calculation of an interference index which is the rel-ative amount of holistic processing for faces in two different carcontexts. A significant correlation of this interference index with ameasure of car expertise can be obtained for several reasons, andour goal was to try to understand what led to this correlation.

The interference index is a difference of differences: the congru-ency effect for faces in the context of normal cars – the congruencyeffect for faces in the context of transformed cars, with each con-gruency effect being a difference score itself. One concern withcorrelations with difference scores is that the variance captured ina correlation can come from the main condition, the control con-dition, or both (e.g., DeGutis et al., 2013). This is not necessarilya problem, depending on the construct measured, but it can leadto misleading interpretations. The difference score that yields thecongruency effects is central to the definition of holistic process-ing as a failure of selective attention and authors generally do notconsider its components further (DeGutis et al., 2013; Richler andGauthier, 2014). In contrast, the difference between holistic pro-cessing in the two contexts is not a unitary construct. The originalprediction is that interference occurs in one context (when bothfaces and cars are shown in their normal configurations) and thatit is not found in the other context (when cars are transformed sothat they do not engage expert processes in car experts).

Here we first replicated the original finding (Experiment 1A),then unpack the effect in ways that were not explored before. Inparticular, we ask whether interference as a function of car exper-tise is attributable to the condition in which faces are shown ina normal car context. To preview our results, we find that it isnot, and so we set out to compare the effect to different baselineconditions, in the hope of clarifying the locus of the effect. InExperiment 1B, we test our prediction that when comparing to abaseline with no irrelevant parts, it would be the car experts’ per-formance that would show interference, and not car novices. Thebaseline will also help characterize the interference as facilitationin congruent trials or interference in incongruent trials. Finally inExperiment 2, we replace cars with novel objects to assess whetherthe interference between two domains can be obtained when per-formance on the interleaved task is matched, but does not tap intoexpert processes.

EXPERIMENT 1ATo assess the robustness of this effect, we first conducted areplication of the study previously reported in Gauthier et al.(2003).

METHODParticipantsThirty-five individuals with normal or corrected-to-normal visionvolunteered to participate for payment: 17 self-reported as carexperts and 18 as novices (six women, one reporting as a carexpert). The rights of the subjects were protected according to aprotocol approved by Vanderbilt University’s Institutional ReviewBoard. The data from two novices were later discarded, onebecause of poor overall performance in the task (54%) and theother because he was an outlier (>3 SD) on our interference index(see design and procedures).

StimuliFor the car expertise test, 120 pictures of different year and/ormodel cars and 120 pictures of different bird species from view-points varying from profile to three-quarter view were used(Figure 1). In the interference task, 336 grayscale (256 × 256pixels) composite images of cars (profile) and faces (front view)made out of the top and bottom of different original images (64faces and 64 cars) were used (Figure 2; see Gauthier et al., 2003).All images had a horizontal red line covering the seam betweenthe two parts. In half of the car images the top part was inverted.The stimuli were presented on a 19-inch monitor with a displayresolution of 1280 × 960 pixels. Participants sat ∼70 cm from thescreen. The position of participants’ heads was not fixed.

Design and procedureSelf-report of expertise is not always a good predictor of perfor-mance (Diamond and Carey, 1986; Rhodes and McLean, 1990;McGugin et al., 2012b) and thus participants were required to per-form a car expertise test (Figure 1; see Gauthier et al., 2000) inaddition to the main interference task (Figure 2). This car exper-tise test yielded a quantitative estimate of their perceptual skillwith cars relative to their skill with a baseline category, birds. Over224 trials, participants matched sequentially presented, 256 × 256-pixel, grayscale images of cars and birds on the basis of their model

Frontiers in Psychology | Cognition September 2014 | Volume 5 | Article 955 | 43




FIGURE 1 | Sequential matching paradigm with images of cars and

birds was used to measure car expertise (relative to a baseline of

performance in bird matching; Gauthier et al., 2000, 2005).

FIGURE 2 | 2-back interleaved part-matching task designed to measure

holistic processing for cars and for faces in a situation where the

processing for both categories overlaps in time (Gauthier et al., 2003).

Composite faces were interleaved with composite cars in either (A) anintact (familiar) or (B) transformed (tops inverted) configuration.

or species (see Figure 1). The first image was presented for 1000 msand was followed by a mask for 500 ms. Then the second imageappeared and remained until either the subject made a responseor 5000 ms had passed. Performance on the bird trials provided abaseline measure for individual differences at subordinate-levelmatching for a category of familiar objects in the absence ofexpertise. As in Gauthier et al. (2003), a car expertise score wascalculated by subtracting the d′ for birds from the d′ for cars foreach individual.

In the interference task, participants performed 1020 trials (60practice, 960 experimental) in which an image was presented cen-trally either for 1500 ms or until they made a response. Imagesalternated between car and face composites (see Figure 2). Par-ticipants pressed a key indicating whether the bottom of thecurrent image was the same or different from the last imageof the same category, triggering the presentation of the nextimage. Thus, participants performed a 2-back part-matching taskin which they were told to always ignore the top of cars andfaces.

Similar to the paradigm used in Gauthier et al. (2003), car con-figuration was manipulated to influence the extent to which theyshould elicit HP in car experts: (i) an upright normal condition(Figure 2A) and (ii) an inverted-top condition (Figure 2B). Thetwo interference conditions alternated in 15 blocks of 60 trials

(a break was given every 30 trials). Half of the trials were con-gruent, where the information from the to-be-ignored top partswould lead subjects to make the same judgment as the informa-tion from the attended bottom part (when compared to the 2-backstimulus from the same category). The other trials were incon-gruent; information from the to-be-ignored top part would leadsubjects to make the opposite judgment as the attended bottompart.

Notably, if participants could follow instructions and com-pletely ignore the top part of composites when making 2-backjudgments on the bottom part, it would make no differencewhether the top part was congruent or incongruent with the cor-rect response for the bottom part. Thus, the degree to which theirrelevant top parts influence judgments about the task-relevantbottom part provides an index of HP (as in Wenger and Ingvalson,2002, 2003; Gauthier et al., 2003).

ANALYSISFace-matching trials performed in the context of cars with invertedtops and those performed in the context of cars with upright topswere split into congruent and incongruent trials. The car trialswere also split into congruent and incongruent trials. Sensitiv-ity (d′) was calculated for the congruent and incongruent trials foreach of the face (upright car-top context, inverted car-top context)and car (upright tops, inverted tops) conditions. HP was opera-tionalized as the sensitivity for congruent minus incongruent trials(HP = d′

congruent – d′incongruent).

The Interference index was then calculated by subtracting theamount of HP for faces in the high interference condition, wherethe faces were processed in the context of upright cars, from that inthe low interference condition, where faces were processed in thecontext of cars with inverted tops. This index provides a measureof the change in HP of the faces due to manipulating the configu-ration of the cars. Because modifying the configuration of objectsof expertise has been shown to impact HP (Young et al., 1987), thisindex will allow us to detect any trade-offs in HP between the twotasks. Crucially the faces and the cars presented in both conditionswere identical except for the orientation of the top (irrelevant) partof the cars, and therefore any difference in HP of the faces betweenthe two conditions can be attributed to the context within whichthe faces were processed.

RESULTSExpertise in car recognition varied from none to extensive. Therewas little variability in bird-matching performance (none of ourparticipants reported any special experience with birds and bird-matching performance was low, ranging from 0.12 to 1.38 d′,consistent with their self-report) compared to car-matching per-formance where d′ scores ranged from 0.37 to 3.76. Consistentwith past work, there was a modest, non-significant, correlationbetween car and bird scores (r32 = 0.28, p = 0.10).

Even though participants were never asked to make a judgmentabout the top, they apparently could not refrain from processingboth faces and cars holistically (see Table 1). This bias was strongerfor faces than cars (t32 = 7.941, p < 0.0001, d = 2.81), this islikely a result of more extensive expertise with faces (Gauthier andTarr, 2002). Normal cars were processed more holistically than





Table 1 | Sensitivity (d ′ and % accuracy) and the derived measures of holistic processing and interference for subjects in Experiment 1A divided

in a novice and expert group according to a median split on the measure of car expertise.

Congruent Incongruent Holistic processing

(congruent–

incongruent)

Car sensitivity

Novices

Inverted car tops 1.57 (77.1) 1.03 (68.8) 0.54

Normal cars 1.39 (74.9) 0.62 (62.0) 0.77

Experts

Inverted car tops 1.90 (81.6) 1.61 (77.6) 0.29

Normal cars 1.93 (82.6) 0.97 (68.0) 0.96

Face sensitivity Interference index (HP new

context–HP old context)

Novices

Transformed car context 1.89 (81.9) 0.76 (64.8) 1.12

Familiar car context 1.80 (80.7) 0.48 (59.5) 1.32

−0.20

Experts

Transformed car context 2.41 (87.3) 0.82 (65.9) 1.58

Familiar car context 2.07 (83.1) 0.63 (62.4) 1.44

0.14

transformed cars, a manipulation check (HP expressed as �d′for normal cars: 0.86 ± 0.07, for transformed cars: 0.42 ± 0.09,t32 = 3.601, p < 0.002, d = 1.27).

Faces seen in the context of normal vs. transformed cars led toapproximately the same degree of HP when expertise was ignored(HP for faces seen in the context of normal cars: 1.38 ± 0.08,for faces seen in the context of transformed cars: 1.35 ± 0.10,t32 = 0.289, p = 0.77, d = 0.10). Critically however, when carexpertise was taken into account, HP for faces depended on theconfiguration of the interleaved cars; as predicted, individuals withhigher levels of car expertise had higher interference indexes (HPfor faces seen in the context of transformed cars minus that forfaces seen in the context of normal cars; r32 = 0.45, F = 7.67,p = 0.009; Figure 3)1.

Intriguingly, while most experts had a positive interferenceindex, suggesting that they processed faces more holistically inthe context of cars in a modified, rather than intact, config-uration, most novices had a negative interference index. Thissuggests that car novices actually processed faces more holisti-cally in the context of cars in an intact, rather than modified,configuration. We also looked at the correlations between carexpertise and each of the two face conditions separately: carexpertise did not predict HP for faces viewed among normal

1Here we subtracted bird scores from car scores to index car expertise, consistentwith prior work with this paradigm. However, in more recent work with car experts,we have regressed out performance in a non-car task from a car task to assessdomain-specific effects (McGugin et al., 2014). The partial correlation between theinterference index and car d′, partialing out bird d′, was r31=0.38, p = 0.01).

FIGURE 3 | Relationship between the participants’ car expertise score

(sensitivity for cars minus their sensitivity for birds in the expertise

test) [�d ′] and the face interference index defined as the amount of

holistic processing for faces when cars were in a new configuration

compared to the holistic processing of faces when cars were in their

normal configuration.

cars (r32 = −0.01, p = 0.96), whereas it predicted HP for facesviewed among transformed cars (r32 = 0.45, p = 0.007), asignificant difference (Steiger’s Z = −2.05, p = 0.04). These find-ings suggest that the interference between faces and cars occurs





in the transformed car condition, which was not the originalprediction.

DISCUSSIONConsistent with previous findings, Experiment 1A provided evi-dence of interference between face and car processing as a functionof expertise with cars (Gauthier et al., 2003). These data demon-strate that interference across different domains of perceptualexpertise, as measured via the impact on an established indexof holistic perception, is a robust and replicable effect. However,the effect was unexpectedly driven by an interaction between facesand transformed cars. That is, car novices and car experts differedmore in their face processing in a transformed car context thanin the familiar car context (see Table 1). This is inconsistent withthe original predictions that motivated this task (i.e., that in carexperts, face recognition would compete most with the processingof whole cars that produce more HP).

One reason why these findings are relatively difficult to inter-pret is that car experts may differ from novices in the amountof HP in several ways. For instance, car experts may show moreHP of cars than car novices because they experience more inter-ference from incongruent to-be-ignored parts, more facilitationfrom to-be-ignored parts, or both. Likewise, the difference in HPof faces by car experts and car novices in the different car con-texts can also be driven by a difference in facilitation, interferenceor both. In addition, while we measured the effect of car exper-tise as a continuous variable, it may seem reasonable to predictthat relative to a baseline condition where holistic processing isnot implicated, it would be performance for the car experts, andnot for the novices, that would show an interaction with faceprocessing. But because our initial prediction that faces and nor-mal cars would be mainly responsible for this interaction was notsupported, we decided to better characterize the effect. Whilethe theoretical significance of whether facilitation and/or inter-ference are critical in this interaction across domains is unclearat the moment, the proposed link between expertise and holisticprocessing makes a strong prediction that novices should not beaffected by the presence of the irrelevant part, while car expertsshould.

Therefore, in Experiment 1B, we tested new car experts andnovices to provide baseline conditions without to-be-ignoredparts. These baselines will be used to estimate whether car expertsand car novices in Experiment 1A differ most in facilitation fromcongruent to-be-ignored parts, or interference from to-be-ignoredparts.

EXPERIMENT 1BTo ask whether the expertise effects observed in Experiment 1Ainfluence facilitation, interference, or both, we presented partici-pants with only the bottom half of each image. This experimentprovided baseline measures to which the sensitivity of congruentand incongruent trials could be compared.

METHODParticipantsTwenty-two (six females) subjects with normal or corrected-to-normal vision and varying levels of car expertise who had

not performed in Experiment 1A participated in this study forpayment or course credit. The rights of the subjects were pro-tected according to a protocol approved by Vanderbilt University’sInstitutional Review Board.

MaterialsStimuli were the same as Experiment 1A except the top half ofeach was removed.

Design and procedureThe procedure was the same as in Experiment 1A except only thebottom half of the car and face images were presented (i.e., theparts above the red line in Figure 2 were omitted).

ANALYSISThe data were divided into expert and novice groups based onparticipants’ car expertise indices (i.e., the difference between car-and bird-sensitivity in the car expertise test; see Experiment 1Afor details). Consistent with previous studies, we defined expertsas individuals with a d′ for cars >2 and a car expertise index(car d′–bird d′) >1 (Gauthier et al., 2000). The data from Exper-iment 1B was also divided into groups of experts and novices.To facilitate comparison across Experiments 1A and 1B, the aver-age expertise index for each of the groups was matched acrossthe two studies. This was done in such a way as to excludedata from as few participants as possible, based only on theircar expertise scores. The resulting mean �d′ and sample sizes(Experiment 1A/Experiment 1B) for the groups were as follows:Expert 1.83 (N = 14)/1.81 (N = 9) and Novice 0.37 (N = 14)/0.36(N = 9).

In general it is more statistically powerful to use car expertise asa continuous variable than as a dichotomous variable. However,because Experiments 1A and 1B include different subjects who arenot matched individually but in groups of novices and experts,we report next a series of ANOVAS on Experiment 1A alone andrelative to the baselines obtained in Experiment 1B in which carexpertise is treated as a dichotomous variable.

RESULTSANOVA comparing expert and novice performance in the 2-back(tops present) task (Experiment 1A)A 2 (group; novice, expert) × 2 (car top context; upright, inverted)was performed on the HP measures for the new groups createdfrom the data reported in Experiment 1A. Consistent with a roleof car expertise in modulating the effect of car context on faceprocessing, as revealed in the correlation analysis reported above,a significant interaction emerged between group and car context,F(1,26) = 5.97, p = 0.022, η2 = 0.064. Car context had a differenteffect on face processing depending on car expertise; for novices,inverting the top of the cars led to a decrease in HP for faces. In con-trast, for car experts, inverting the top of the cars led to an increasein HP for faces. There was also a marginally significant effect ofgroup, F(1,26) = 4.13, p = 0.053, η2 = 0.090, suggesting that carexperts in general processed the faces more holistically. There wasno main effect of car context on HP of faces (F < 1). Plannedt-tests showed that while HP for faces among normal cars failedto differentiate between car experts and novices, t(26) = 0.30,p = 0.77, d = 0.12, car experts processed faces shown among





transformed cars more holistically than car novices, t(26) = 2.76,p = 0.01, d = 0.48.

Planned comparisons on HP for cars revealed more HPfor upright cars than cars with inverted tops in car experts,t(26) = 3.58, p = 0.001, d = 1.40, but not in novices, t(26) = 1.03,p = 0.31, d = 0.40 (see Figures 4A,B). In fact, HP was signif-icantly different from 0 in all car conditions (all ps ≤ 0.0005)except for cars with inverted tops in car experts (p = 0.13). Thissuggests that inverting car tops made them easier to ignore for carexperts.

ANOVA comparing expert and novice performance in 2-back half(tops absent) task (Experiment 1B)A 2 (group; experts, novices) × 2 (category; faces, cars) ANOVAon the results for the baseline task (no-top) revealed a signifi-cant interaction between group and category, F(1,16) = 7.34,p = 016, η2 = 0.041. As expected, car novices and experts didnot differ in their performance matching face bottoms (p > 0.05),but car experts were better at matching car bottoms than novices(p < 0.001).

Comparison of Experiment 1A against baselines fromExperiments 1BFigure 4 illustrates the results in Experiment 1A in each condition,for car novices and car experts, relative to the baselines obtain inExperiments 1B where no to-be-ignored parts were present. As wehave seen, these baselines indicate when participants performedthe task with only the to-be-attended parts, car novices and carexperts only differed in their processing of car parts, being betterin car experts. Now we use these baselines to interpret the resultsof Experiment 1A and specifically ask in what way car novices andcar experts differ.

The results of planned t-tests comparing sensitivity (d′) in thetop-present congruent and incongruent conditions (from Experi-ment 1A) with their respective no-top baselines (from Experiment1B) reveals that HP during the top-present task was mainly due tointerference from incongruent top halves rather than facilitationfrom congruent ones (Figure 4). Incongruent conditions led tolower sensitivity than their respective baselines (all ps < 0.05) in allcases except for cars with inverted tops in both groups (ps < 0.05).In contrast, the only condition with significant facilitation was for

FIGURE 4 | Sensitivity for car, (A,B), and face, (C,D), matching judgments

of bottom halves of composites for groups of car novices, (A,C), and

experts, (B,D), in Experiment 1A. In each graph, the results from the twoconditions in Experiment 1A (when the top of cars were upright and when

they were inverted) are plotted with the baseline provided by matched groupsof novices and experts in Experiment 1B, matching the bottom halves offaces or cars with no top half present. Error bars show the standard error ofthe mean.





faces seen in the context of cars with inverted tops, in car experts(p < 0.05).

Summary and discussionCar experts experienced less interference than car novices fortransformed cars (Figures 4A vs. B), while they experiences morefacilitation from congruent face tops processed in this context(Figures 4C vs. D).

One account of these results is that because car experts pro-cessed cars in a transformed configuration less holistically thanregular cars, they therefore recruited less HP resources. Therefore,once the top part of the car is inverted it can be effectively ignoredby car experts.

Because the car context was manipulated, while the face taskwas the same in both car contexts, and because car novices andcar experts did not differ in their processing of face parts in a carcontext when there was no to–be-ignored part, it is reasonable toinfer that this is what led to facilitation by congruent face parts inthe transformed car context. Exactly why this happened though isunclear but it is plausible that the transformed car context allowedcar experts to reduce executive control when car tops were easierto ignore, and as a result also exerted less effort trying to ignoreface tops.

Most experiments using the composite task have not used abaseline condition to examine facilitation/interference, but whenit has been used, the congruency effect obtained for aligned faceswas generally due to interference from incongruent parts, withoutsignificant facilitation (Richler et al., 2008). This is consistent withwhat we observed here in car novices (even if the baseline wasobtained in a different group). This highlights how abnormal theprocessing of faces among transformed cars was in our car expertparticipants.

It should be noted that the choice of a baseline condition such asthe isolated relevant parts used in Experiment 1B can be difficult.In the comparison between Experiments 1A and 1B, there are dif-ferences in stimuli (parts vs. composites) and in task requirements(requirement to selectively attend or not).

Given the complexity of the 2-back interleaved dual task, itis possible that other task components unrelated to HP andaffected by car expertise, such as more general effects related tothe executive control and/or short-term memory load demandsof the task, play a role in producing these effects. Experiment2 investigates an alternative account of the interference betweendifferent expert domains, assessing whether explanations appeal-ing to contributions from these more general effects can be ruledout.

EXPERIMENT 2Our proposed account of why car experts showed more facilita-tion from to-be-ignored congruent face parts in the context oftransformed cars points to how car experts were better able toignore inverted car tops. Experts showed no congruency effectsof inverted car tops while novices did. Non-face objects are notprocessed holistically in the composite paradigm, which gener-ally means that they do not show more of a congruency effect ina normal than transformed (typically misaligned) configuration.However, there is sometimes a small but significant congruency

effect for non-face objects that is not modulated by configuration(e.g., Wong et al., 2009; Richler et al., 2011b). There are situationswhere training has a main effect of reducing this congruency effectfor stimuli in a transformed configuration (Chua et al., 2014),similar to the difference between our car novices and experts inExperiment 1A.

The absence of holistic processing for cars with invertedtops among car experts, but not novices, suggests an alterna-tive account of the interference between face and car processingthat is unrelated to expertise. Because it was the transformedcar context that drove the interference effect, it is possible thatthe same interference effect would be observed when faces areprocessed in the context of any object category that does notshow a congruency effect, regardless of expertise. This wouldsuggest that an alternative account, grounded in the more gen-eral demands of the two concurrently performed tasks, ratherthan in participants’ expertise, would better explain this interfer-ence.

In the original paradigm, the condition where car novices pro-cess faces in the context of normal cars should had offered a testof this hypothesis. However, there was a significant congruencyeffect for cars in car novices here, perhaps in part because they hadsome non-negligible experience with cars (this congruency effectcould also be amplified by cars being processed in the context offaces, see Richler et al., 2009).

Therefore, to test this hypothesis, we designed simple and unfa-miliar stimuli in an oval shape with parts defined by coloredgratings (varying in both hue and luminance, see Figure 5A) andin a pilot experiment, we verified that their processing in the com-posite task did not produce any congruency effect. Here we askwhether faces processed in our dual task with these “egg” stimuliwould produce the same facilitation as observed in Experiment1A. If the increase in the facilitation component of HP for facesprocessed in the context of transformed cars is simply due to thefact that the to-be-ignored parts of these transformed objects areeasy to ignore, then we should observe it here. In contrast, iffacilitation is not observed, this would indicate that the interac-tion between selective attention to faces and the car context isspecifically dependent on car expertise.

METHODParticipantsWe recruited fourteen volunteer participants (seven females), tomatch the size of the car expert group in Experiment 1. The rightsof the subjects were protected according to a protocol approved byVanderbilt University’s Institutional Review Board.

StimuliFace stimuli were the same as in Experiment 1. Stimuli were con-structed from 64 oval shapes or “eggs” approximately the samesize as the face stimuli with 15 vertical stripes of two alternat-ing colors. Twelve different colors were used, selected from theAdobe Photoshop© palette to be similar but still distinguishable(all shades of blue, green or purple). The 64 original eggs weresplit in half and recombined (in the same manner as the facestimuli) to create 128 composite eggs used in the experiment (seeFigure 5A).





FIGURE 5 | Results for Experiment 2. (A) Examples of egg stimuli, this pair has the same bottom but incongruent tops. (B) Sensitivity for egg and facejudgments in Experiment 2. The results of congruent and incongruent trials are plotted against the baseline obtained from the same subjects performing thesame task with no top half present.

Design and procedureThe procedure was identical to that of Experiment 1, except thatwhole faces were processed in only one context, that is the orien-tation of the top part of the egg was not manipulated (Figure 5A).The same participants also performed a baseline task with no topson faces or eggs (as in Experiment 1B). There were an equal num-ber of top-present and no-top trials, presented in four blocks of240 trials. The two conditions alternated and their order was coun-terbalanced across participants. As in Experiment 1, participantsperformed 2-back judgments on the bottom half of all images.

RESULTS AND DISCUSSIONAs expected, the eggs produced a pattern of sensitivity very similarto that of cars with an inverted top in Experiment 1 (Figure 5B).There was no significant HP for eggs, t(13) = 1.21, p = 0.25,d = 0.67. Sensitivity for the egg baseline did not differ from thebaseline for cars among car experts in Experiment 1, t(21) = 0.77,p = 0.45, d = 0.34. Sensitivity for eggs was also not significantlydifferent from that for cars with inverted tops among experts inExperiment 1, both in the congruent, t(26) = 0.68, p = 0.50,d = 0.27, and incongruent conditions, t(26) = 1.16, p = 0.26,d = 0.45.

However, unlike in Experiment 1, this context did not leadto facilitation for congruent face trials. There was no signifi-cant facilitation (congruent trials relative to no-top baseline) forfaces processed in the context of eggs, t(13) = 1.97, p = 0.07,d = 1.09. Comparing with Experiment 1, the estimate of facil-itation from faces among eggs was both indistinguishable fromthat obtained from car novices matching faces among transformedcars, t(26) = 1.27, p = 0.22, d = 0.50, and significantly less thanthat obtained from car experts matching faces among cars withinverted tops, t(26) = 2.39, p = 0.02, d = 0.94.

The results of Experiment 2 suggest that it is not the lack ofHP from the transformed car context per se that led to facilitationfrom congruent face trials in Experiment 1 as the egg stimuli werealso not processed holistically, yet did not impact holistic faceperception. Further, the findings of Experiment 1 also demonstratethat the interferences was not simply a general effect of participantscar expertise (because in the upright car condition there was nofacilitation from the to-be-ignored part for faces) nor becausefaces were processed among cars (because facilitation for faceswas not obtained in car novices). Rather, face HP was specificallyinfluenced by the concurrent processing of transformed objects ofexpertise.





GENERAL DISCUSSIONOur results replicated the interference between holistic process-ing of faces and cars observed in Gauthier et al. (2003). Havingfound that the interaction depended on the processing of facesamong transformed cars, we investigated these effects furtherby using a dual-task with isolated parts to partition congruencyeffects into interference from incongruent parts and facilitationfrom congruent parts. We found that car expertise was associ-ated with less interference from incongruent car parts, and thatin this context, congruent face parts produced more facilitation.Finally, we showed that this facilitation effect for faces was notobtained in another dual task with objects that produced nocongruency effect, suggesting that the interaction depends onexpertise.

Our results highlight the fact that transformed objects of exper-tise can lead to effects that are distinct from control objects. Whilethere has been much more focus on how the processing of wholeobjects of expertise is special (both faces and objects), there is noquestion that experts also process parts differently. This is mostclearly shown in our results by the advantage of car experts oncar novices for car part performance. There are other examplesof transformed objects of expertise producing effects that are dis-tinct from control objects. For example, the ERP response (N170)to inverted faces is larger than the response for upright faces anddelayed by 10 ms, while other objects elicit a response of muchsmaller amplitude and invariant to orientation (Rossion et al.,2000). This is also found in subjects who have been trained withnon-face objects (Rossion et al., 2002). It is possible that theseresponses triggered by transformed objects of expertise index amechanism that can interfere with expert processing of objectsfrom another category.

Recent findings from studies of other domains of expertise,such as chess, that have also been shown to result in increased HP asindexed via the composite task, support the suggestion that trans-formed stimuli of expertise can trigger stronger responses in brainregions linked with perceptual expertise than their intact versions.For example, greater FFA activation was found in response to chessstimuli among chess experts when the structure of the chess stim-uli was distorted compared to intact (Bilalic et al., 2011). Further,recent findings suggest that disrupted objects of expertise, such asthe transformed cars used here, may trigger a search for structureor meaningful chunks by experts that appears to also involve afrontal-parietal network (Bartlett et al., 2013; Rennig et al., 2013;see also Bor and Owen, 2007). These existing findings, and thefrequency with which transformed objects of experts (inverted,scrambled, misaligned etc.) are used as a control or compari-son stimulus category, highlight the importance of further studiesexploring the processing of such objects among experts.

Another possibility is that the expertise-related interferencebetween face and car processing primarily reflects an attention-based effect. For instance, because car experts can more easilyselectively attend to the bottom part and thus ignore the (task-irrelevant) top part in the transformed car condition, they may“relax” control of their attention in this task-context, resultingin more intrusions of congruent face parts. The effect could becarried by facilitation because interference from incongruent faceparts may already be strong to start with. Our egg control task

suggests that such facilitation is not observed in any situationwhere the to-be-ignored part is easily ignored – perhaps a cer-tain degree of fluency with the relevant parts is also required.This account is obviously quite speculative and will require furthertesting.

Although this study cannot reveal where in the brain inter-actions between faces and cars may occur, our task is likely toengage parietal and frontal areas implicated in short-term mem-ory (Goldman-Rakic, 1987; Belger et al., 1998) as well as theFFA (Courtney et al., 1997; Grady et al., 1998; Haxby et al., 2000;Druzgal and D’Esposito, 2001), which is the part of the brain mostassociated with the idea of a “face module” (Kanwisher et al., 1997;McCarthy et al., 1997). Activity in the FFA increases directly withthe short-term memory load for faces (Druzgal and D’Esposito,2001). This region is a plausible candidate for a locus of inter-ference obtained in dual experts for a number of reasons. Inparticular, it is recruited for both cars and faces in car experts(Gauthier et al., 2000, 2005; McGugin et al., 2012a); the activityin this area in response to cars correlates with behavioral mea-sures of car expertise (Gauthier et al., 2000) and although it isnot the only area to show an effect of expertise for cars, whenthe task demands are made more difficult, as was the case here,effects in these other regions drop whereas the expertise effectin the FFA remains (McGugin et al., in press). Finally, a recentfMRI study found that car expertise effects in the FFA survivedmanipulations of clutter and of divided attention, but that theywere abolished when cars were presented in the context of faces,especially when the faces were also task-relevant (McGugin et al.,2014). Much work remains to be done to relate this exampleof competition between faces and cars in the FFA with wholeobjects and the present finding of interaction between faces andtransformed objects of expertise during a dual-task that requiresselective attention.

Other findings of expertise-dependent interference between theconcurrent processing of face and non-face objects of expertisesuggest that not only can faces and objects of expertise be pro-cessed in a similar manner, as well as neurally close in space andtime, but the neural networks responsible for their HP may notbe functionally independent (Gauthier et al., 2003; Rossion et al.,2004, 2007; McKeeff et al., 2010; McGugin et al., 2011). Impor-tantly, it is not necessary to postulate that processing of faces andcars depend on overlapping sets of neurons to account for compe-tition between the two domains. It is sufficient to assume that faceand object processing are closer together in experts than novices in“cerebral functional space”(Kinsbourne and Hicks, 1978); cerebralfunctional space refers to the physical size of and distance betweenbrain areas responsible for different functions. This only assumesthat competition is more likely between neural ensembles that aremore densely interconnected, and/or that are separated by fewersynapses.

In conclusion, our results are consistent with previous findingsof observable interference across different domains of real-worldexpertise where the particular domains are proposed to rely ona common resource. This functional overlap between face andnon-face domains of expertise has implications for the potentialof extensive learning, as in the case of real-world expertise, to leadto a dynamic reorganization of cognitive resources. Because most





normal adults possess a certain degree of expertise with faces,it may be important to consider training and application envi-ronments for real-world experts and assess the extent to whichcompetition between different domains can impact learning andperformance.

ACKNOWLEDGMENTSKim M. Curby was supported by funding from the AustralianResearch Council (DE130100969). This work was also supportedby funding from NSF Award BCS-0091752, National Eye InstituteGrant R01-EY13441, and a grant from the James S. McDonnellFoundation to Isabel Gauthier.

REFERENCESBahrick, H. P., Bahrick, P. O., and Wittlinger, R. P. (1975). Fifty years of memory for

names and faces: a cross-sectional approach. J. Exp. Psychol. Gen. 104, 54–75. doi:10.1037/0096-3445.104.1.54

Bartlett, J. C., Boggan, A. L., and Krawczyk, D. C. (2013). Expertise andprocessing distorted structure in chess. Front. Hum. Neurosci. 7:825. doi:10.3389/fnhum.2013.00825

Belger, A., Puce, A., Krystal, J. H., Gore, J. C., Goldman-Rakic, P., and McCarthy,G. (1998). Dissociation of mnemonic and perceptual processes during spatialand nonspatial working memory using fMRI. Hum. Brain Mapp. 6, 14–32. doi:10.1002/(SICI)1097-0193(1998)6:1<14::AID-HBM2>3.0.CO;2-O

Bilalic, M., Langner, R., Ulrich, R., and Grodd, W. (2011). Many faces of expertise:fusiform face area in chess experts and novices. J. Neurosci. 31, 10206–10214. doi:10.1523/JNEUROSCI.5727-10.2011

Boggan, A. L., Bartlett, J. C., and Krawczyk, D. C. (2012). Chess masters show ahallmark of face processing with chess. J. Exp. Psychol. Gen. 141, 37–42. doi:10.1037/a0024236

Bor, D., and Owen, A. M. (2007). Cognitive training: neural correlates of expertskills. Curr. Biol. 17, R95–R97. doi: 10.1016/j.cub.2007.01.019

Bruce, V., Doyle, T., Dench, N., and Burton, M. (1991). Rememberingfacial configurations. Cognition 38, 109–144. doi: 10.1016/0010-0277(91)90049-A

Bukach, C. M., Phillips, W. S., and Gauthier, I. (2010). Limits of generalizationbetween categories and implications for theories of category specificity. Atten.Percept. Psychophys. 72, 1865–1874. doi: 10.3758/APP.72.7.1865

Chua, K., Richler, J., and Gauthier, I. (2014). Becoming a Lunari or Taiyo expert:learned attention to parts drives holistic processing of faces. J. Exp. Psychol. Hum.Percept. Perform. 40, 1174–1182. doi: 10.1037/a0035895

Courtney, S. M., Ungerleider, L. G., Keil, K., and Haxby, J. V. (1997). Transient andsustained activity in a distributed neural system for human working memory.Nature 386, 608–611. doi: 10.1038/386608a0

Curby, K. M., and Gauthier, I. (2007). A visual short-term memory advantage forfaces. Psychon. Bull. Rev. 14, 620–628. doi: 10.3758/BF03196811

Curby, K. M., Glazek, K., and Gauthier, I. (2009). A visual short-term memoryadvantage for objects of expertise. J. Exp. Psychol. Hum. Percept. Perform. 35,94–107. doi: 10.1037/0096-1523.35.1.94

Curby, K. M., Goldstein, R. R., and Blacker, K. (2013). Disrupting perceptual group-ing of face parts impairs holistic face processing. Atten. Percept. Psychophys. 75,83–91, doi: 10.3758/s13414-012-0386-9

DeGutis, J., Wilmer, J., Mercado, R. J., and Cohan, S. (2013). Using regression tomeasure holistic face processing reveals a strong link with face recognition ability.Cognition 126, 87–100. doi: 10.1016/j.cognition.2012.09.004

Dehaene, S., and Cohen, L. (2011). The unique role of the visual word form area inreading. Trends Cogn. Sci. 15, 254–262. doi: 10.1016/j.tics.2011.04.003

Diamond, R., and Carey, S. (1977). Developmental changes in the represen-tation of faces. J. Exp. Child Psychol. 23, 1–22. doi: 10.1016/0022-0965(77)90069-8

Diamond, R., and Carey, S. (1986). Why faces are and are not special: an effectof expertise. J. Exp. Psychol. Gen. 115, 107–117. doi: 10.1037/0096-3445.115.2.107

Druzgal, T. J., and D’Esposito, M. (2001). Activity in fusiform face area modu-lated as a function of working memory load. Cogn. Brain Res. 10, 355–364. doi:10.1016/S0926-6410(00)00056-2

Dundas, E. M., Plaut, D. C., and Behrmann, M. (2013). The joint development ofhemispheric lateralization for words and faces. J. Exp. Psychol. Gen. 142, 348–358.doi: 10.1037/a0029503

Farah, M. J., Wilson, K. D., Drain, M., and Tanaka, J. N. (1998). What is“special” about face perception? Psychol. Rev. 105, 482–498. doi: 10.1037/0033-295X.105.3.482

Gauthier, I., Curby, K. M., Skudlarski, P., and Epstein, R. A. (2005). Indi-vidual differences in FFA activity suggest independent processing at differentspatial scales. Cogn. Affect. Behav. Neurosci. 5, 222–234. doi: 10.3758/CABN.5.2.222

Gauthier, I., Curran, T., Curby, K. M., and Collins, D. (2003). Perceptual interferencesupports a non-modular account of face processing. Nat. Neurosci. 6, 428–432.doi: 10.1038/nn1029

Gauthier, I., Skudlarski, P., Gore, J. C., and Anderson, A. W. (2000). Expertise forcars and birds recruits brain areas involved in face recognition. Nat. Neurosci. 3,191–197. doi: 10.1038/72140

Gauthier, I., and Tarr, M. J. (2002). Unraveling mechanisms for expert objectrecognition: bridging brain activity and behavior. J. Exp. Psychol. Hum. Percept.Perform. 28, 431–446. doi: 10.1037/0096-1523.28.2.431

Goldman-Rakic, P. S. (1987). “Circuitry of primate prefrontal cortex and regulationof behavior by representational memory,” in Handbook of Physiology: The NervousSystem, Vol. 5, ed. F. Plum (Bethesda, MD: American Physiological Society).

Grady, C. L., McIntosh, A. R., Bookstein, F., Horwitz, B., Rapoport, S. I., andHaxby, J. V. (1998). Age-related changes in regional cerebral blood flow dur-ing working memory for faces. Neuroimage 8, 409–425. doi: 10.1006/nimg.1998.0376

Haig, N. D. (1984). The effect of feature displacement on face recognition. Perception13, 505–512. doi: 10.1068/p130505

Haxby, J. V., Petit, L., Ungerleider, L. G., and Courtney, S. M. (2000). Distinguishingthe functional roles of multiple regions in distributed neural systems for visualworking memory. Neuroimage 11, 380–391. doi: 10.1006/nimg.2000.0592

Hosie, J. A., Ellis, H. D., and Haig, N. D. (1988). The effect of feature displacementon the perception of well-known faces. Perception 17, 461–474. doi: 10.1068/p170461

Kanwisher, N. (2000). Domain specificity in face perception. Nat. Neurosci. 3,759–763. doi: 10.1038/77664

Kanwisher, N., McDermott, J., and Chun, M. M. (1997). The fusiform face area: amodule in human extrastriate cortex specialized for face perception. J. Neurosci.17, 4302–4311.

Kemp, R., McManus, C., and Pigott, T. (1990). Sensitivity to the displacementof facial features in negative and inverted images. Perception 19, 531–543. doi:10.1068/p190531

Kinsbourne, M., and Hicks, R. E. (1978). “Functional cerebral space: a modelfor overflow, transfer and interference effects in human performance: a tutorialreview,” in Attention and Performance VII, ed. J. Requin (Hillsdale, NJ: LawrenceErlbaum Associates), 345–362.

McCarthy, G., Puce, A., Gore, J. C., and Allison, T. (1997). Face-specific pro-cessing in the human fusiform gyrus. J. Cogn. Neurosci. 9, 605–610. doi:10.1162/jocn.1997.9.5.605

McGugin, R. W., Gatenby, J. C., Gore, J. C., and Gauthier, I. (2012a). High-resolutionimaging of expertise reveals reliable object selectivity in the fusiform face arearelated to perceptual performance. Proc. Natl. Acad. Sci. U.S.A. 109, 17063–17068.doi: 10.1073/pnas.1116333109

McGugin, R. W., Richler, J. J., Herzmann, G., Speegle, M., and Gauthier, I. (2012b).The Vanderbilt Expertise Test reveals domain-general and domain-specific sexeffects in object recognition. Vision Res. 69, 10–22. doi: 10.1016/j.visres.2012.07.014

McGugin, R. W., and Gauthier, I. (2010). Perceptual expertise with objects predictsanother hallmark of face perception. J. Vis. 10, 15.1–15.12. doi: 10.1167/10.4.15

McGugin, R. W., McKeeff, T. J., Tong, F., and Gauthier, I. (2011). Irrelevant objectsof expertise compete with faces during visual search. Atten. Percept. Psychophys.73, 309–317. doi: 10.3758/s13414-010-0006-5

McGugin, R. W., Newton, A. T., Gore, J. C., and Gauthier, I. (in press). Robustexpertise effects in right FFA. Neuropsychologia.

McGugin, R. W., Van Gulick, A. E., Tamber-Rosenau, B. J., Ross, D. A., and Gauthier,I. (2014). Expertise effects in face-selective areas are robust to clutter and divertedattention, but not to competition. Cereb. Cortex doi: 10.1093/cercor/bhu060[Epub ahead of print].





McKeeff, T. J., McGugin, R. W., Tong, F., and Gauthier, I. (2010). Expertise increasesthe functional overlap between face and object perception. Cognition 117, 355–360. doi: 10.1016/j.cognition.2010.09.002

Rennig, J., Bilalic, M., Huberle, E., Karnath, H.-O., and Himmelbach, M.(2013). The temporo-parietal junction contributes to global gestalt perception –evidence from studies in chess experts. Front. Hum. Neurosci. 7:513. doi:10.3389/fnhum.2013.00513

Rhodes, G., and McLean, I. G. (1990). Distinctiveness and expertise effects withhomogeneous stimuli: towards a model of configural coding. Perception 19, 773–794. doi: 10.1068/p190773

Richler, J. J., Bukach, C. M., and Gauthier, I. (2009). Context influences holisticprocessing of nonface objects in the composite task. Percept. Psychophys. 71,530–540. doi: 10.3758/APP.71.3.530

Richler, J. J., Cheung, O. S., and Gauthier, I. (2011a). Holistic processing predictsface recognition. Psychol. Sci. 22, 464–471. doi: 10.1177/0956797611401753

Richler, J. J., Mack, M. L., Palmeri, T. J., and Gauthier, I. (2011b). Invertedfaces are (eventually) processed holistically. Vision Res. 51, 333–342. doi:10.1016/j.visres.2010.11.014

Richler, J. J., and Gauthier, I. (2014). A meta-analysis and review of holistic faceprocessing. Psychol. Bull. doi: 10.1037/a0037004 [Epub ahead of print].

Richler, J. J., Palmeri, T. J., and Gauthier, I. (2012). Meanings, mechanisms, andmeasures of holistic processing. Front. Psychol. 3:553. doi: 10.3389/fpsyg.2012.00553

Richler, J. J., Tanaka, J. W., Brown, D. D., and Gauthier, I. (2008). Why does selectiveattention to parts fail in face processing? J. Exp. Psychol. Learn. Mem. Cogn. 34,1356–1368. doi: 10.1037/a0013080

Rossion, B., Collins, D., Goffaux, V., and Curran, T. (2007). Long-term expertisewith artificial objects increases visual competition with early face catego-rization processes. J. Cogn. Neurosci. 19, 543–555. doi: 10.1162/jocn.2007.19.3.543

Rossion, B., Gauthier, I., Goffaux, V., Tarr, M. J., and Crommelinck, M. (2002).Expertise training with novel objects leads to left lateralized face-like electro-physiological responses. Psychol. Sci. 13, 250–257. doi: 10.1111/1467-9280.00446

Rossion, B., Gauthier, I., Tarr, M. J., Despland, P., Bruyer, R., Linotte, S., et al. (2000).The N170 occipito-temporal component is delayed and enhanced to invertedfaces but not to inverted objects: an electrophysiological account of face-specificprocesses in the human brain. Neuroreport 11, 69–74. doi: 10.1097/00001756-200001170-00014

Rossion, B., Kung, C. C., and Tarr, M. J. (2004). Visual expertise with nonface objectsleads to competition with the early perceptual processing of faces in the humanoccipitotemporal cortex. Proc. Natl. Acad. Sci. U.S.A. 101, 14521–14526. doi:10.1073/pnas.0405613101

Simons, D. J. (2014). The value of direct replication. Perspect. Psychol. Sci. 9, 76–80.doi: 10.1177/1745691613514755

Tanaka, J. W. (2001). The entry point of face recognition: evidence for face expertise.J. Exp. Psychol. Gen. 130, 534–543. doi: 10.1037/0096-3445.130.3.534

Tanaka, J. W., and Taylor, M. (1991). Object categories and expertise: is the basiclevel in the eye of the beholder? Cogn. Psychol. 23, 457–482. doi: 10.1016/0010-0285(91)90016-H

Wenger, M. J., and Ingvalson, E. M. (2002). A decisional component of holisticencoding. J. Exp. Psychol. Learn. Mem. Cogn. 28, 872–892. doi: 10.1037/0278-7393.28.5.872

Wenger, M. J., and Ingvalson, E. M. (2003). Preserving informational separabilityand violating decisional separability in facial perception and recognition. J. Exp.Psychol. Learn. Mem. Cogn. 29, 1106–1118. doi: 10.1037/0278-7393.29.6.1106

Wong, A. C.-N., Palmeri, T. J., and Gauthier, I. (2009). Conditions for face-likeexpertise with objects: becoming a Ziggerin expert – but which type? Psychol. Sci.20, 1108–1117. doi: 10.1111/j.1467-9280.2009.02430.x

Young, A. W., Hellawell, D., and Hay, D. (1987). Configural information in faceperception. Perception 10, 747–759. doi: 10.1068/p160747


Received: 30 June 2014; accepted: 11 August 2014; published online: 10 September2014.Citation: Curby KM and Gauthier I (2014) Interference between face and non-facedomains of perceptual expertise: a replication and extension. Front. Psychol. 5:955.doi: 10.3389/fpsyg.2014.00955This article was submitted to Cognition, a section of the journal Frontiers in Psychology.Copyright © 2014 Curby and Gauthier. This is an open-access article distributed underthe terms of the Creative Commons Attribution License (CC BY). The use, distributionor reproduction in other forums is permitted, provided the original author(s) or licensorare credited and that the original publication in this journal is cited, in accordance withaccepted academic practice. No use, distribution or reproduction is permitted whichdoes not comply with these terms.







doi: 10.3389/fpsyg.2014.00793

Visual appearance interacts with conceptual knowledge inobject recognitionOlivia S. Cheung1,2* and Isabel Gauthier3

1 Department of Psychology, Harvard University, Cambridge, MA, USA2 Center for Mind/Brain Sciences, University of Trento, Trentino, Italy3 Department of Psychology, Vanderbilt University, Nashville, TN, USA

Edited by:


Reviewed by:

Elan Barenholtz, Florida AtlanticUniversity, USAThomas A. Farmer, University ofIowa, USA

*Correspondence:

Olivia S. Cheung, Department ofPsychology, Harvard University, 33Kirkland Street, Cambridge, MA02138, USAe-mail: [email protected]

Objects contain rich visual and conceptual information, but do these two types ofinformation interact? Here, we examine whether visual and conceptual informationinteract when observers see novel objects for the first time. We then address howthis interaction influences the acquisition of perceptual expertise. We used two typesof novel objects (Greebles), designed to resemble either animals or tools, and two lists ofwords, which described non-visual attributes of people or man-made objects. Participantsfirst judged if a word was more suitable for describing people or objects while ignoringa task-irrelevant image, and showed faster responses if the words and the unfamiliarobjects were congruent in terms of animacy (e.g., animal-like objects with words thatdescribed human). Participants then learned to associate objects and words that wereeither congruent or not in animacy, before receiving expertise training to rapidly individuatethe objects. Congruent pairing of visual and conceptual information facilitated observers’ability to become a perceptual expert, as revealed in a matching task that required visualidentification at the basic or subordinate levels. Taken together, these findings show thatvisual and conceptual information interact at multiple levels in object recognition.

Keywords: object learning, semantics, visual features, perceptual expertise

INTRODUCTIONA chocolate bunny is more visually similar to a stuffed animal butmore conceptually similar to a baking chocolate bar, and the com-bination is such that a child may not allow her parent to melt itto bake a cake, nor would the parent allow the child to bring it inbed. Our interactions with objects must take both visual and con-ceptual information into account but little research addresses howobject recognition mechanisms are constrained by the interactionsbetween these two sources of information.

Object perception involves more than processing visual fea-tures. For familiar objects, visual knowledge, such as color ofa fruit, modulates perception of salient features of an object(Hansen et al., 2006; Witzel et al., 2011). Conceptual knowl-edge about familiar object categories is also represented in thevisual system (e.g., animals, tools, Chao et al., 1999; Mahon andCaramazza, 2009; Huth et al., 2012). While is often assumedthat visual features of novel objects engage minimal concep-tual processing (Tarr and Pinker, 1989; Bülthoff and Edelman,1992; Gauthier and Tarr, 1997; Hayward and Williams, 2000;Schwoebel and Srinivas, 2000; Curby et al., 2004; Bar and Neta,2006; Op de Beeck et al., 2006), shape dimensions of novel objects(e.g., sharpness, symmetry, contrast, complexity) can impactobservers’ subjective preferences (Reber et al., 2004; Bar and Neta,2006, 2007). Moreover, intuitions may also be formulated aboutthe similarity of novel objects to familiar objects (e.g., smoothnovel objects resembling “women wearing hats,” Op de Beecket al., 2006, p.13031), and such meaningful interpretations of

ambiguous shapes appear to be robust and stable within indi-vidual observers (Voss et al., 2012). However, how meaningsevoked by visual features may influence object processing remainsa question that has not been explored systematically.

Some information on how object representations are con-strained by both visual and conceptual factors comes from exper-iments where new conceptual associations are created for visualstimuli. Conceptual associations can facilitate perceptual catego-rization (Wisniewski and Medin, 1994; Lin and Murphy, 1997),bias perceptual interpretation of neutral stimuli (Bentin andGolland, 2002; Hillar and Kemp, 2008), and improve visual dis-crimination (Dux and Coltheart, 2005; Lupyan and Spivey, 2008).The discriminability of shapes or faces increases after having beenpaired with words from different categories, compare with hav-ing been paired with words from similar categories (Dixon et al.,1997, 1998; Gauthier et al., 2003). Observers also activate recentconceptual associations during visual judgments, even when theinformation is task irrelevant (James and Gauthier, 2003, 2004).However, in these studies (e.g., Dixon et al., 1997, 1998; Gauthieret al., 2003; James and Gauthier, 2003, 2004), the conceptual andvisual information are arbitrarily associated, leaving entirely openwhether some of these associations are created more easily thanothers, such as when the visual and conceptual features conveycongruent, compared to contradictory, information.

We start with the assumption that the animate/inanimate dis-tinction exists in the visual arena (objects can look like an animalor not) as well as in the non-visual conceptual arena (we can list













Cheung and Gauthier Visual-conceptual interaction in object recognition

attributes of objects that are animate or not). In this study, wemanipulated both visual and conceptual features to study theirinteraction, more specifically the alignment of an animate vs.inanimate dimension in the visual and conceptual domains. Weused words that described non-visual attributes that would nor-mally apply to either people or man-made objects (e.g., cheerful,affordable), and created novel objects that resembled either liv-ing or non-living things. For visual features, we attempted toconvey the animate vs. inanimate character of novel objects bymanipulating shape, texture and color. These dimensions werechosen because bilateral shape symmetry is a powerful indica-tor of animacy (Concar, 1995; Rhodes et al., 1998), whereas theshape of man-made objects is more variable depending on theirfunction. Also, the objects were rendered in colors and texturesgenerally associated with animals or tools (e.g., skin color/organicvs. non-skin color/metallic). Experimental manipulation of bothconceptual and visual information afforded us more control toinvestigate their interaction.

INITIAL VISUAL-CONCEPTUAL BIASESWe first examined to what extent the visual appearance of novelobjects from unfamiliar categories evokes conceptual processing,when observers see the objects for the first time. We asked whethervisual features of the “animal-like” and “tool-like” object setsare sufficient to evoke the conceptual biases of animacy. Insteadof asking participants directly to categorize the novel objects asanimate or inanimate entities, we tested if the visual appear-ance of the objects evoked the concepts related to animate vs.inanimate categories by testing whether their (task-irrelevant)presence interfered with judgments of non-visual attributes asbeing more relevant to people or to man-made objects (e.g.,“excited,” “grateful” vs. “durable,” “useful”).

VISUAL-CONCEPTUAL INTERACTION ON EXPERT RECOGNITIONBeyond any early conceptual biases evoked by visual appear-ance, it is also possible that visual-conceptual interactions becomemore important with experience with a category. If visual featuresof novel objects activate abstract biases, anchoring the objects intoexisting conceptual networks appropriately (e.g., calling animate-like objects “animals” vs. calling tool-like objects “animals”) mayconstrain their representations during expertise training. Theremay be differences in the acquisition of expertise between objectsthat look like animals or not (i.e., the effect of visual appear-ance), or between objects that are introduced as having animateor inanimate conceptual properties (i.e., the effect of concep-tual associations). But more importantly, we asked whether it iseasier to acquire expertise with a category that is assigned concep-tual features congruent with its appearance (i.e., the interactionbetween visual and conceptual information), as we conjecturedthat learning objects with congruent visual and conceptual infor-mation might enhance the ability to locate diagnostic visualfeatures for fine-level discrimination.

TRAINING PROCEDURESHere we combined training procedures used in previous con-ceptual association studies (James and Gauthier, 2003, 2004)and expertise studies (Gauthier and Tarr, 1997, 2002; Wong

et al., 2009). During the two-stage training, participants firstlearned to associate particular concepts with individual objects,and then learned to rapidly recognize objects at the subordinatelevel. Critically, participants were divided into two groups dur-ing the first training stage: Both groups were shown identicalwords and objects, but the Congruent pairing group learned toassociate animate attributes with animal-like objects and inan-imate attributes with tool-like objects, while the Incongruentpairing group learned the opposite pairings. In the second train-ing stage, both groups practiced individuating objects from bothanimal-like and tool-like categories, without further mention ofconceptual information.

DEPENDENT MEASURESWe used two dependent measures to reveal potential visual-conceptual interactions. First, in a word judgment task, partici-pants categorized words as appropriate for describing people orman-made objects presented on task-irrelevant objects. This taskwas first completed prior to any training, and then completedafter each training stage. This task uses an opposition logic simi-lar to the Stroop task (1935) and several tasks since (e.g., see Bubet al., 2008), to test whether the visual appearance of the animal-like and tool-like objects would be sufficient to evoke conceptsrelevant to animacy/non-animacy. If our manipulation of visualappearance does not evoke animate vs. inanimate concepts, wordjudgment performance should not be affected by whether con-gruent or incongruent objects are present. While the actual locusof any interference may be at the response level, such responseswould have to be evoked by visual appearance (note that at pre-test, no response had ever been associated with these or similarobjects).

Second, in an object matching task, participants judged if twoobjects were from the same category (basic-level trials), or showedthe same individual (subordinate-level trials). The reduction ofthe “basic-level advantage” is a hallmark of real-world expertise(Tanaka and Taylor, 1991), which is also sensitive to short-termexpertise training (Bukach et al., 2012). Expert observers recog-nize individual objects in their expert categories at the subordinatelevel (e.g., “eastern screech owl,” or “Tom Hanks”) as quickly asat the basic level (e.g., “bird,” or “man”), whereas novices recog-nize the objects faster at the basic than the subordinate levels (i.e.,the “basic-level advantage,” Rosch et al., 1976). The basic-leveladvantage is reduced in experts for both animate and inani-mate object categories (e.g., faces: Tanaka, 2001; birds: Tanakaet al., 2005; Scott et al., 2006, 2008; novel 3D objects: Gauthieret al., 1998; Wong et al., 2009; Wong et al., 2012). With novelobjects, explicit conceptual information is often absent duringtraining (e.g., Wong et al., 2009; Wong et al., 2012). Althoughfaster subordinate-level processing in experts might depend pre-dominantly on experience with perceptual information of similarexemplars in a category, it is possible that conceptual informationalso impose processing constraints. For instance, brief learning ofa diverse set of semantic associations with novel objects can facil-itate subordinate-level judgment compared to that of a restrictedset (Gauthier et al., 2003). The question of interest here is whetherobservers apply conceptual knowledge about familiar categoriesto novel objects, based on the visual resemblance between the






familiar and novel categories. If this is the case, conceptual infor-mation expected based on past experience with visual featuresmay facilitate fine-level discrimination of similar exemplars, asboth visual and conceptual information interact to constrainobject representations.

Here, we assessed whether having associated concepts thatare congruent with the visual appearance of a category mayfacilitate the recognition of the objects at the subordinate-levelcompared to the basic-level, even though the object matchingtask can be accomplished based on visual features alone. We mea-sured any differences in the “basic-level advantage” after semantictraining and after individuation training. If visual processingis facilitated by visual-conceptual pairings, then the basic-leveladvantage should be more reduced in participants who associatedthe objects with congruent conceptual features than in those whoreceived incongruent pairings.

METHODSPARTICIPANTSTwenty-four adults (normal/corrected-to-normal vision) fromVanderbilt University participated for payment ($12/h). Thestudy was approved by the Vanderbilt University IRB. Participantswere randomly assigned to the Congruent pairing group (6females and 6 males, age M = 22.58, SD = 4.32) or theIncongruent pairing group (4 females and 8 males, age M =23.67, SD = 4.29). Twelve additional adults (5 females and7 males, age M = 22.67, SD = 3.08) participated only in theobject-matching task once as a Control group.

STIMULIObjectsEach participant was shown 48 novel objects called “Greebles”(see examples in Figure 1A) created using 3D Studio Max. Halfof the objects (24) were Symmetric-organic Greebles with smooth-edged parts and organic textures. The rest (24) were Asymmetric-metallic Greebles with sharp-edge parts and metallic textures. Notethat symmetry refers to object, and not image, symmetry. EachGreeble had a unique set of four peripheral parts. To minimizeobject-specific effects, we generated two versions of Symmetric-organic and Asymmetric-metallic Greebles that differed in color(i.e., yellow/pink, blue/green), central and peripheral part assign-ment to the objects. Each version was shown to half of theparticipants in each of the two training groups. There were 18Greebles from each of the Symmetric-organic and Asymmetric-metallic categories in the trained subsets, and 6 in the untrainedsubsets (which were used as foils in the basic-level recognitiontask). The two subsets (trained or untrained) within each cate-gory had different central and peripheral parts. From each trainedsubset, six Greebles were used in semantic training. An additionalsix Greebles from each trained subset were also used in individ-uation training. All objects were shown during the testing tasks.The objects used for training and testing were counterbalancedacross participants within each group and matched betweengroups. All Greebles were rendered on a white background at fourviewpoints (0/6/12/18◦: The 0◦ view was an arbitrarily definedorientation with the symmetric axis rotated 40◦ to the right). Theimage size was approximately 6 × 3.6◦ of visual angle. To avoid

FIGURE 1 | (A) Examples of the two categories of objects and twocategories of words, including Symmetric-organic objects,Asymmetric-metallic objects, animate attributes and inanimate attributes.(B) Schematic of the two-stage training. In semantic training (stage 1),participants were divided into two groups to learn to associate three wordsto each trained object. The two training groups differed only in terms of thepairing of the objects and words. In individuation training (stage 2), allparticipants learned to name and identify objects at the subordinate levelquickly and accurately.

image-based effects, objects used during training were shown at 0and 18◦. During testing, the objects were presented at 6 and 12◦.Additionally, phase-scrambled images of the Greebles were alsocreated as control stimuli in one of the tasks.

WordsEighty-four words were used; each described a non-visualattribute appropriate for describing either people (“animateattributes”) or man-made objects (“inanimate attributes”;Figure 1A and Appendix A in Supplementary Material ), gener-ated in a pilot study (N = 20). Word length was controlled acrossthe animate (M = 7.17 letters, SD = 2.05) vs. inanimate (M =7.5 letters, SD = 1.86) features. According to the SUBTLEXus

word frequency database (Brysbaert and New, 2009), the meanfrequency was higher for the animate (M = 38.06, SD = 63.53)than inanimate (M = 4.16, SD = 6.29) attributes. But sincethe critical manipulation here was the object-word pairing andidentical words were used for both training groups, word fre-quency alone could not account for differences between groups.Twenty-four animate and 24 inanimate attributes were used inthe word judgment task. Eighteen animate and 18 inanimateattributes were used during semantic training. The words used





were also counterbalanced across participants within each groupand matched between groups.

PROCEDUREThe study was conducted on Mac mini computers with 19′′CRT monitors using Matlab. Below are the details on the two-stage training (which lasted approximately 9 h across 6 days),the word judgment and object matching tasks (which lasted 15and 45 min respectively). The entire study consisted of a pre-test(word judgment task), followed by two sessions of semantic train-ing, followed by a session with two post-test tasks (word judgmentand object matching), followed by four sessions of individuationtraining, then another session with the two post-test tasks.

TrainingTraining stage 1: semantic training. During semantic train-ing (two 90-min sessions; Figure 1B, Table 1), each group

learned three randomly selected words each for 12 Greebles(6 Symmetric-organic Greebles and 6 Asymmetric-metallicGreebles). The Congruent pairing group learned animateattributes with Symmetric-organic Greebles and inanimateattributes with Asymmetric-metallic Greebles, whereas theIncongruent pairing group learned the opposite pairing. Identicalsets of word triplets were assigned to one participant in theCongruent pairing group and another in the Incongruent pairinggroup. The two categories of Greebles were shown in interleavedblocks.

Training stage 2: individuation training. During individua-tion training (four 90-min sessions; Figure 1B, Table 1), allparticipants learned to individuate 24 Greebles (12 Symmetric-organic and 12 Asymmetric-metallic Greebles; in which 6 fromeach category were previously shown during semantic training).Additional objects were used in this phase to increase the

Table 1 | Task details of the two-stage training paradigms.

SEMANTIC TRAINING

Session and number of objects involved No. of trials Task

Session 1 (4 Symmetric-organic Greeblesand 4 Asymmetric-metallic Greebles),Session 2 (6 Symmetric-organic Greeblesand 6 Asymmetric-metallic Greebles)

16 in session 1,

24 in session 2

Passive viewing: To initiate learning, this task allowed participants to studyeach Greeble with the three associated attributes, twice for as long asneeded

576 Three-attribute matching: To promote associations between each Greebleand each unique set of attributes, this task required participants to judge if aset of three attributes matched a concurrently presented Greeble

576 Single-attribute matching: To ensure participants learned all three attributesindependently, instead of any one from each set, this task requiredparticipants to judge if a single attribute matched a subsequently presentedGreeble

16 in session 1,24 in session 2

Recall: To examine if participants were able to generate the associatedattributes without verbal hints, participants were asked to input the threeattributes associated with each Greeble, twice

INDIVIDUATION TRAINING

Session and number of objects involved No. of trials Task

Session 1 (6 Symmetric-organic Greeblesand 6 Asymmetric-metallic Greebles),Session 2 (12 Symmetric-organic Greeblesand 12 Asymmetric-metallic Greebles)

720 Naming: To promote learning of each Greeble with its name, participantswere asked to input the first letter of the name associated with a Greeble.The names were shown during the first 3 presentations of a Greeble

480 Name matching: To ensure participants learn to individuate the Greeblesquickly and accurately, participants were asked to judge if a name matchedwith a concurrently presented Greeble as quickly and accurately as possible

384 Name verification: A variation of the Name matching task to encouragetask-general learning, participants judged if a name matched with asubsequently presented Greeble

We used a variety of tasks in every session to promote task-general learning. These tasks were previously used in several studies (semantic training: James and

Gauthier, 2003, 2004; individuation training: e.g., Gauthier and Tarr, 2002; Wong et al., 2009). The semantic training (stage 1) consisted of four tasks promoting

associations between a set of three words to a trained object. The trained objects were introduced across the first two training sessions: 8 Greebles (4 Symmetric-

organic and 4 Asymmetric-metallic Greebles) were introduced in session 1 and all 12 Greebles (6 of each category) were introduced in session 2. The individuation

training (stage 2) consisted of three tasks that aimed to enhance the speed and accuracy of identification for individual objects at the subordinate level. The trained

objects were introduced across the first two training sessions: 12 Greebles were used in session 1 and all 24 Greebles were used in sessions 2–4.






difficulty of rapid identification. During this training, eachGreeble was named with a 2-syllable nonsense word (e.g., Pila,Aklo, see Appendix B in Supplementary Material for the full list).Name assignment was randomized within group but matchedbetween groups. Both speed and accuracy were emphasized inall training tasks. To motivate participants, the mean speedand accuracy for each block were shown at the end of eachblock. Symmetric-organic and Asymmetric-metallic Greebleswere shown in interleaved blocks.

TestingWord judgment. Participants first completed the 15-min task(Figure 2A) prior to training, and again after semantic train-ing and after individuation training. In this 2-alternative forcedchoice task, participants judged if a word was more appro-priate for describing people or objects, while told to ignorean image presented behind each word. In a total of 432 tri-als, each of the 24 animate and 24 inanimate attributes waspresented nine times. Each word appeared twice with each ofthe 48 Symmetric-organic Greebles and 48 Asymmetric-metallicGreebles at each of two slightly different viewpoints (difference= 6◦), and four times with each of 12 phase-scrambled Greebleimages (6 Symmetric-organic and 6 Asymmetric-metallic). Thephase-scrambled images were included to evaluate whether par-ticipants paid additional attention to the task-irrelevant Greeblesduring the word judgment task. All stimuli were shown until aresponse, with a 1-s interval in between trials. All conditions wererandomized.

FIGURE 2 | Example trials of (A) the word judgment task and (B) the

object matching task at the basic level (top: a basic-level trial with two

individuals from different categories) and at the subordinate level

(bottom: a subordinate-level trial with two individuals from the same

category).

Matching at basic- and subordinate-levelsParticipants completed this 45-min task (768 trials) after semantictraining and after individuation training (Figure 2B). In differentblocks, participants judged if two sequentially presented objectswere identical or different, at either the basic or subordinate level.In basic-level blocks, object pairs could be Greebles from the samecategory (the same central body part) or different categories (dif-ferent central body parts). In subordinate-level blocks, the objectpairs could be identical or different individuals from the same cat-egory (the same central body parts but different peripheral parts).All object pairs were shown across 6◦ rotation. The following con-ditions were blocked: Categorization level (basic/subordinate),Visual appearance (Symmetric/Asymmetric), and training status(trained/untrained objects). On each trial, a 300 ms-fixation wasfollowed by a study image (800 ms), a mask (500 ms), and by atest image (1 s).

RESULTSTRAINING RESULTSThe training was meant to form conceptual associations andimprove individuation performance, and the training results(Figure 3) were not a focus of the study. Both groups showedaccuracy near ceiling throughout training (i.e., well above 90% inall tasks across all sessions), with the expected significant increasesin all individuation training tasks. Responses became faster withtime in all semantic training and individual training tasks butthe single-attribute matching task. Note that responses were alsofaster for Symmetric-organic Greebles than Asymmetric-metallicGreebles, but there was no statistical significant difference inperformance between the groups in all but the passive viewingtask during semantic training. We do not report statistical analy-ses here, but Figure 3 shows confidence intervals relevant to thesignificant training effects across sessions.

TESTING RESULTSWord judgmentWe focused on RT in correct trials because accuracy in this taskwas high (>95%)1.

There was no effect of the Pairing group on this task at anystage of the study (pre-test, after semantic training or individua-tion training) ANOVAs (all p > 0.35). There results are thereforepresented in Figure 4 collapsing over this factor.

It was entirely expected that there would be no differencebetween the two pairing groups at pre-test because no pairingshad actually been done. At this stage, the question was whetherthe visual appearance of novel objects imply conceptual informa-tion about animacy. We also measured performance in the wordjudgment in a baseline condition where task-irrelevant scrambledimages were shown behind the words (Table 2).

As mentioned above, we observed no effect of Pairing groupsafter pairings were learned in semantic training. Like any nullresult, this is difficult to interpret, but given we found othereffects of pairing group in the study (described later), this suggeststhat the word judgment is simply not sensitive to these effects.

1Trials that involved the word “curvy” were discarded because of chanceperformance across participants.





FIGURE 3 | Mean response times in the various tasks of (A)

semantic training and (B) individuation training. Note that RT wasin seconds in the Passive viewing task and in milliseconds in all

other tasks. Error bars represent the 95% confidence intervals of thetraining effects across sessions within each group for each objecttype.

FIGURE 4 | Mean response times (ms) in the word judgment task as a

function of Word type (animate attributes vs. inanimate attributes) and

Object type (Symmetric-organic Greebles vs. Asymmetric-metallic

Greebles) (A) in the first session (pre-training), (B) in the second session

(post-semantic training), and (C) in the third session (post-individuation

training). Error bars represent the 95% confidence intervals of the Wordtype and Object type interaction. The lines were added to the figures despitethe conditions being categorical, to highlight the interactions.

This could be because it is an explicit conceptual task in whichparticipants can as easily retrieve all the explicitly learned asso-ciations, whether congruent or incongruent. In contrast, in amore perceptual task where no explicit conceptual search is acti-vated, congruent visual-conceptual pairings may show more of anadvantage.

Interaction between visual appearance and conceptual infor-mation. After collapsing over the non-significant factor of

pairing group, we focus here on the effect of Visual appear-ance on word categorization across sessions (Figure 4). ASession (pre-training/post-semantic training/post-individuationtraining) × Word type (animate/inanimate) × Object type(Symmetric-organic/Asymmetric-metallic) ANOVA was con-ducted. Responses became faster with time, F(2, 46) = 13.53,η2

p = 0.37, p < 0.0001. Responses were also faster for judging

animate than inanimate attributes, F(1, 23) = 10.60, η2p = 0.32,

p = 0.0035, possibly because of the higher word frequency for






Table 2 | Mean response times (ms) in the word judgment task for

each image type (Symmetric-organic Greebles, Asymmetric-metallic

Greebles, and scrambled images) across the three testing sessions.

Symmetric-organic Asymmetric-metallic Scrambled

Greebles Greebles images

Session 1 798.9 (25.2) 805.7 (24.5) 830.9 (28.7)

Session 2 758.0 (29.0) 759.3 (28.4) 768.2 (27.3)

Session 3 707.5 (21.4) 709.5 (21.2) 723.2 (19.5)

Standard errors of the mean are reported within the parentheses.

animate than inanimate attributes. There was no significant effectof Object type, F(1, 23) = 0.86, η2

p = 0.04, p = 0.36, nor signifi-cant interactions between Session and Word type, F(2, 46) = 0.48,η2

p = 0.02, p = 0.62, or between Session and Visual appearance,

F(2, 46) = 0.35, η2p = 0.015, p = 0.71. Critically, however, Word

type and Object type interacted, F(1, 23) = 6.68, η2p = 0.25, p =

0.017, although the 3-way interaction of Session, Word type andObject type did not reach significance, F(2, 46) = 2.31, η2

p = 0.09,p = 0.11.

One of our main goals was to investigate whether Word typeand Visual appearance might interact when participants were firstpresented with the objects during the pre-training sessions, andhow a putative interaction would be affected by further training.Therefore, we conducted a Word type (animate/inanimate) ×Object type (Symmetric-organic/Asymmetric-metallic) ANOVAseparately for each session to examine if the effect was alreadysignificant at pre-test (Figures 4A–C). Critically, we found asignificant interaction between Word type and Object type inpre-training [F(1, 23) = 7.18, η2

p = 0.24, p = 0.013]. Scheffé’spost-hoc tests revealed faster judgment for animate attributeswith the presence of Symmetric-organic Greebles compared withAsymmetric-metallic Greebles (p = 0.0045), and a significanteffect of the opposite result for judging inanimate attributes(p = 0.015).

The Word type and Object type interaction was also signifi-cant after individuation training [F(1, 23) = 6.16, η2

p = 0.21, p =0.02], but interestingly, it was not immediately after semantictraining [F(1, 23) = 0.033, η2

p = 0.0015, p = 0.86]. We observe abias for relating novel animal-like (or tool-like) objects to human(or object) attributes at pre-test, and it seems that introducingexplicit semantic associations can temporarily alter this bias. Thisis also consistent with the idea, suggested above to explain thelack of a Pairing group effect, that this explicit word judgmenttask may be most sensitive to implicit influences. During seman-tic training, participants in all groups had to learn associationswith the objects, and the training ensured that all associationswere learned. These explicit associations would have been moresalient to the minds of participants in Session 2 than later on.We would therefore speculate that these associations blocked theeffects of visual appearance that we observe in Sessions 1 and 3,and the reappearance of the interaction effect in Session 3 demon-strates that the faster RTs or practice with the word judgment taskcannot account for the lack of effect in Session 2.

Manipulation check: objects vs. scrambled objects as task-irrelevant images. To test whether participants paid less attention

to the words during the word judgment task due to the presenceof task-irrelevant objects, we compared performance to that forthe same task with words shown on scrambled images. The pres-ence of an object was apparently not more distracting than thepresence of a scrambled image, in fact if anything the objects wereeasier to ignore than the scrambled images (perhaps due to low-level image properties). Indeed, RTs for the word judgment wereconsistently faster when objects were present relative to scram-bled images (Table 2). A Session (pre-training/post-semantictraining/post-individuation training) × Image type (Symmetric-organic/Asymmetric-metallic/Scrambled) ANOVA showed aneffect of Image type, F(2, 46) = 8.26, η2

p = 0.26, p < 0.001, withfaster RT with the presence of either type of objects comparedto the scrambled images (ps < 0.01), and no difference betweenobject types (p = 0.80). There was also an effect of Session,F(2, 46) = 15.53, η2

p = 0.40, p < 0.0001, with faster RT as the ses-sions progressed, and no interaction between Session and Imagetype, F(4, 92) = 1.11, η2

p = 0.05, p = 0.37.

Matching at the basic- and subordinate-levelsAfter finding that the visual appearance of novel objects can acti-vate conceptual information in a word judgment task on the firstencounter with these objects, we then examined the influenceof acquired conceptual associations with animate vs. inanimateobjects, in a matching task at the basic- and subordinate-levels.As in prior work (e.g., Gauthier and Tarr, 1997; Wong et al., 2009;Wong et al., 2011), we focus only on trials with unfamiliar objectsfrom the trained categories that were not used during training(i.e., “transfer” objects)2, as a critical aspect of expertise is general-ization of the skills to unfamiliar exemplars in the expert domain(e.g., car experts viewing cars, Bukach et al., 2010, face expertsviewing faces, Tanaka, 2001). Here, we measured both responsetimes (RT) and sensitivity (d′: z(hit rate)-z(false alarm rate)).We first compared the performance of the two training groupsafter semantic training to an untrained control group. We thencompared the effects in the two training groups after both stages(semantic and individuation) of training.

Effects of semantic training (comparison between a Controlgroup and the training groups). Semantic training with afew exemplars was sufficient to reduce basic-level advantage,even for untrained exemplars in the training groups com-pared to a Control group that did not receive any train-ing (Figures 5A,B). A Group (Control/Congruent/Incongruent)× Object type (Symmetric-organic/Asymmetric-metallic) ×Categorization level (Basic/Subordinate) ANOVA revealed a sig-nificant interaction of Group and Categorization level in RT,F(1, 33) = 9.00, η2

p = 0.40, p < 0.001: the basic-level advantagewas smaller in the training groups compared to the control group(ps < 0.05), and also smaller in the Congruent than Incongruentpairing group (p = 0.04). The Group and Categorization level

2For the trained objects, the two training groups showed comparableimprovement in this task after individuation training, with comparable mag-nitude of reduction in basic-level advantage. These results were consistentwith the results from the individuation training procedures whereby bothgroups successfully learned to quickly and accurately identify this subset ofthe exemplars at the subordinate-level





FIGURE 5 | Results of the object matching task. Left panel: Performanceof the Control group who did not receive any training. Right panel:Performance of the two Training groups. Panel (A) shows response times(ms) and Panel (B) shows sensitivity (d′) as a function of Group/Pairing(Control vs. Congruent pairing vs. Incongruent pairing), Session (no training

vs. post-semantic vs. post-individuation training), Visual appearance(Symmetric-organic Greebles vs. Asymmetric-metallic Greebles) andCategorization level (Basic vs. Subordinate). The dots represent meanresponse times or mean sensitivity, and the bars represent the meanbasic-level advantage. Error bars represent the standard errors of the mean.

interaction did not reach significance d′, F(1, 33) = 2.72, η2p =

0.12, p = 0.08.

Effects of both semantic and individuation training (com-parison between the training groups). We then assessed howpairing during semantic training influenced the acquisition ofperceptual expertise in the two training groups. RT and d′were analyzed in a Pairing (Congruent/Incongruent) × Session(post-semantic/post-individuation) × Object type (Symmetric-organic/Asymmetric-metallic: each category was paired witheither animate or inanimate attributes) × Categorization level(Basic/Subordinate) ANOVA, respectively. RT and d′ results(Figures 5A,B) revealed different aspects of conceptual influ-ences: RT showed a long-lasting pairing effect throughout thetests, whereas d′ showed an effect of conceptual association typeonly after semantic training.

In RT, object matching was faster after individuation train-ing than after semantic training, F(1, 22) = 41.17, η2

p = 0.65,p < 0.0001. The basic-level advantage was present, F(1, 22) =109.6, η2

p = 0.83, p < 0.0001, with faster recognition at thebasic level compared to the subordinate level. The basic-level advantage was smaller for Symmetric-organic Greeblesthan Asymmetric-metallic Greebles, F(1, 22) = 12.82, η2

p = 0.37,p = 0.002. Critically, the Congruent pairing group showed areduced basic-level advantage compared to the Reversed pair-ing group, as revealed by an interaction between Pairing

and Categorization level, F(1, 22) = 7.12, η2p = 0.24, p = 0.014.

The interaction of Pairing, Category level, and Session wasnot significant, F(1, 22) = 0.11, η2

p = 0.005, p = 0.74, nor wasany other effect (ps > 0.31). Thus, visual-conceptual pair-ing impacted both matching performance and a markerof perceptual expertise: associations with congruent concep-tual facilitated perceptual judgments, relative to incongruentassociations.

The basic-level advantage was also present in d′, F(1, 22) =75.35, η2

p = 0.77, p < 0.0001. All other results were not signif-icant (all p > 0.09) except for an unexpected result regardingthe type of conceptual associations. This was a 4-way interac-tion of Group, Session, Object type and Categorization level,F(1, 22) = 6.69, η2

p = 0.23, p = 0.017. Although a 4-way interac-tion could be difficult to interpret, the result essentially revealedthat immediately after semantic training, both groups showeda smaller basic-level advantage for Greeble categories associatedwith inanimate attributes compared with the categories associ-ated with animate attributes (ps < 0.006). However, followingindividuation training the basic-level advantage no longer dif-fered depending on animate or inanimate associations (ps >

0.32). Unlike the effect of visual-conceptual pairing in RT thatwas observed both after semantic and individuation training, thetype of conceptual associations had an initial impact on matching,but the effect was absent once the conceptual associations were nolonger emphasized.






DISCUSSIONWe found an implicit bias to relate animate concepts to unfa-miliar symmetric, animal-like objects, and to relate inanimateconcepts to unfamiliar asymmetric, tool-like objects. This is a rareand important experimental demonstration that the processing ofnovel objects is far from neutral conceptually. Moreover, whethervisual and conceptual information are associated in a congruentor incongruent manner influences visual processing of untrainedobjects from the category. These effects last long after associationsare no longer task-relevant.

Consistent with previous work, we found that concepts can bequickly associated with novel objects (Dixon et al., 1997; Gauthieret al., 2003; James and Gauthier, 2003, 2004), and that learningdistinctive semantic associations can facilitate subordinate-levelprocessing (Gauthier et al., 2003). Our results also led us tospeculate that such conceptual associations, especially right afterthey were freshly learned, may in some tasks block the auto-matic activation of semantic information evoked by the visualfeatures themselves. This conjecture is based on the absence of aninteraction between semantics and visual appearance in the wordjudgment task only in Session 2.

For the first time we considered the effect of different kinds ofpairings of conceptual information with novel objects, informa-tion that was either congruent or incongruent with the animacyof the visual appearance. We found that both congruent andincongruent pairings of objects and concepts can be learned.Moreover, these associations generalize to an object category, asthey influenced performance for untrained objects during a visualmatching task.

Specifically, congruent visual-conceptual pairings facilitatedthe acquisition of subordinate-level perceptual expertise, result-ing in a smaller basic-level advantage in the Congruent thanIncongruent pairing group. When learning to individuate objects,observers not only utilize visual information, they are affectedby conceptual cues implied from visual features. The new asso-ciations introduced during semantic training interacted withthe initial conceptual biases for the objects, such that con-gruent cues from different sources facilitate forming preciserepresentations for visually similar exemplars in the trainedcategories.

On the other hand, the fact that even relatively unexpectedconceptual associations (e.g., inanimate attributes to animal-likeobjects) generalized to objects that shared only some of the visualproperties of the trained objects suggests a mechanism to explainthe implicit bias observed in the word judgment task for novelobjects prior to any training. We showed that unfamiliar objectsfrom a novel category (e.g., symmetric-organic objects) appear toderive conceptual meaning on the basis of visual similarity withfamiliar categories (e.g., animals or people). Likewise, unfamiliarobjects from recently familiarized categories (i.e., the untrainedobjects in the trained categories in the current study) derive con-ceptual meaning on the basis of visual similarity to objects from arecently learned category. If relatively novel and arbitrary associa-tions that run contrary to much of our experience can generalizein this manner, a lifetime’s history of conceptual learning likelyhas a very powerful influence on how we represent any object weencounter.

Additionally, while the main focus of the study is on theinteraction between visual and conceptual properties, we foundtransient effects regarding the type of conceptual informationon object processing immediately after associations were learned.For instance, objects associated with inanimate attributes showedless of a basic-level advantage compared to objects associ-ated with animate attributes. One possibility is that inanimateconcepts possess lower feature overlap than animate concepts(Mechelli et al., 2006). Two objects that are “elastic, shinyand antique” vs. “eco-friendly, plastic and durable” may seemto be quite different and likely to belong to different basic-level categories. Conversely, two objects that are “adorable,funny and sensitive” and “cheerful, talented and forgiving” aremore likely two individuals within the same basic-level cate-gory. Therefore, inanimate associations may be more distinc-tive than animate associations, facilitating visual discrimination(Gauthier et al., 2003). Note, however, this difference cannotaccount for the pairing effect, because identical sets of asso-ciations were used for both training groups. Also, this effectregarding the type of associations faded once the associationswere no longer emphasized, even though the visual-conceptualpairing effects remained. Further research should aim to repli-cate and explore the different temporal dynamics of the moreshort-lived effect of distinctive conceptual associations, and thecongruency of the visual-conceptual associations, which werelonger-lasting.

Several influential object recognition theories focus almostentirely on visual attributes of objects (e.g., Marr, 1982;Biederman, 1987; Perrett and Oram, 1993; Riesenhuber andPoggio, 1999; Jiang et al., 2007), assuming that conceptualassociations should have no influence on object recognition(e.g., Pylyshyn, 1999; but see Goldstone and Barsalou, 1998).Additionally, researchers interested in the role of shape in objectprocessing have often used novel objects to prevent influencesfrom non-visual information, such as object names, familiar-ity and conceptual content (e.g., Op de Beeck et al., 2008).Our findings suggest that novel objects are not necessarily con-ceptually neutral, and that both visual and conceptual factors,and their interaction are important in the formation of objectrepresentations.

ACKNOWLEDGMENTSThis research was supported by the James S. McDonnellFoundation, NIH grant 2 R01 EY013441-06A2 and NSF grantSBE-0542013. We thank Timothy McNamara, James Tanaka,Frank Tong and Jennifer Richler for valuable comments on a pre-vious version of the manuscript, and Magen Speegle and SarahMuller for assistance with data collection.


REFERENCESBar, M., and Neta, M. (2006). Humans prefer curved visual objects. Psychol. Sci. 17,

645–648. doi: 10.1111/j.1467-9280.2006.01759.x







Bar, M., and Neta, M. (2007). Visual elements of subjective preference modu-late amygdala activation. Neuropsychologia 45, 2191–2200. doi: 10.1016/j.neuropsychologia.2007.03.008

Bentin, S., and Golland, Y. (2002). Meaningful processing of meaningless stim-uli: the influence of perceptual experience on early visual processing of faces.Cognition 86, B1–B14. doi: 10.1016/S0010-0277(02)00124-5

Biederman, I. (1987). Recognition-by-components: a theory of human imageunderstanding. Psychol. Rev. 94, 115–147. doi: 10.1037/0033-295X.94.2.115

Brysbaert, M., and New, B. (2009). Moving beyond Kucera and Francis: a criticalevaluation of current word frequency norms and the introduction of a new andimproved word frequency measure for American English. Behav. Res. Methods41, 977–990. doi: 10.3758/BRM.41.4.977

Bub, D. N., Masson, M. E. J., and Cree, G. S. (2008). Evocation of functional andvolumetric gestural knowledge by objects and words. Cognition 106, 27–58. doi:10.1016/j.cognition.2006.12.010

Bukach, C. M., Phillips, W. S., and Gauthier, I. (2010). Limits of generalizationbetween categories and implications for theories of category specificity. Atten.Percept. Psychophys. 72, 1865–1874. doi: 10.3758/APP.72.7.1865

Bukach, C. M., Vickery, T. J., Kinka, D., and Gauthier, I. (2012). Training experts:individuation without naming is worth it. J. Exp. Psychol. Hum. Percept. Perform.38, 14–17. doi: 10.1037/a0025610

Bülthoff, H. H., and Edelman, S. (1992). Psychophysical support for a 2-D viewinterpolation theory of object recognition. Proc. Natl. Acad. Sci. U.S.A. 89,60–64. doi: 10.1073/pnas.89.1.60

Chao, L. L., Haxby, J. V., and Martin, A. (1999). Attribute-based neural substratesin temporal cortex for perceiving and knowing about objects. Nat. Neurosci. 2,913–919. doi: 10.1038/13217

Concar, D. (1995). Sex and the symmetrical body. New Sci. 146, 40–44.Curby, K. M., Hayward, W. G., and Gauthier, I. (2004). Laterality effects in the

recognition of depth rotated novel objects. Cogn. Affect. Behav. Neurosci. 4,100–111. doi: 10.3758/CABN.4.1.100

Dixon, M., Bub, D., and Arguin, M. (1997). The interaction of object form andobject meaning in the identification performance of a patient with category-specific visual agnosia. Cogn. Neuropsychol. 14, 1085–1130. doi: 10.1080/026432997381286

Dixon, M., Bub, D., and Arguin, M. (1998). Semantic and visual determinants offace recognition in a prosopagnosic patient. J. Cogn. Neurosci. 10, 362–376. doi:10.1162/089892998562799

Dux, P. E., and Coltheart, V. (2005). The meaning of the mask matters: evidence ofconceptual interference in the attentional blink. Psychol. Sci. 16, 775–779. doi:10.1111/j.1467-9280.2005.01613.x

Gauthier, I., James, T. W., Curby, K. M., and Tarr, M. J. (2003). The influence of con-ceptual knowledge on visual discrimination. Cogn. Neuropsychol. 20, 507–523.doi: 10.1080/02643290244000275

Gauthier, I., and Tarr, M. J. (1997). Becoming a “Greeble” expert: exploring mech-anisms for face recognition. Vision Res. 37, 1673–1682. doi: 10.1016/S0042-6989(96)00286-6

Gauthier, I., and Tarr, M. J. (2002). Unraveling mechanisms for expert objectrecognition: bridging brain activity and behavior. J. Exp. Psychol. Hum. Percept.Perform. 28, 431–446. doi: 10.1037/0096-1523.28.2.431

Gauthier, I., Williams, P., Tarr, M. J., and Tanaka, J. W. (1998). Training “Greeble”experts: a framework for studying expert object recognition processes. VisionRes. 38, 2401–2428. doi: 10.1016/S0042-6989(97)00442-2

Goldstone, R. L., and Barsalou, L. W. (1998). Reuniting perception and conception.Cognition 65, 231–262. doi: 10.1016/S0010-0277(97)00047-4

Hansen, T., Olkkonen, M., Walter, S., and Gegenfurtner, K. R. (2006). Memorymodulates color appearance. Nat. Neurosci. 9, 1367–1368. doi: 10.1038/nn1794

Hayward, W. G., and Williams, P. (2000). Viewpoint dependence and objectdiscriminability. Psychol. Sci. 11, 7–12. doi: 10.1111/1467-9280.00207

Hillar, K. F., and Kemp, R. (2008). Barack Obama or Barry Dunham? The appear-ance of multiracial faces is affected by the names assigned to them. Perception37, 1605–1608. doi: 10.1068/p6255

Huth, A. G., Nishimoto, S., Vu, A. T., and Gallant, J. L. (2012). A contin-uous semantic space describes the representation of thousands of objectand action categories across the human brain. Neuron 76, 1210–1224. doi:10.1016/j.neuron.2012.10.014

James, T. W., and Gauthier, I. (2003). Auditory and action semantic features acti-vate sensory-specific perceptual brain regions. Curr. Biol. 13, 1792–1796. doi:10.1016/j.cub.2003.09.039

James, T. W., and Gauthier, I. (2004). Brain areas engaged during visual judgmentsby involuntary access to novel semantic information. Vision Res. 44, 429–439.doi: 10.1016/j.visres.2003.10.004

Jiang, X., Bradley, E., Rini, R. A., Zeffiro, T., Vanmeter, J., and Riesenhuber,M. (2007). Categorization training results in shape- and category-selectivehuman neural plasticity. Neuron 53, 891–903. doi: 10.1016/j.neuron.2007.02.015

Lin, E. L., and Murphy, G. L. (1997). Effects of background knowledge on objectcategorization and part detection. J. Exp. Psychol. Hum. Percept. Perform. 23,1153–1169. doi: 10.1037/0096-1523.23.4.1153

Lupyan, G., and Spivey, M. J. (2008). Perceptual processing is facilitated by ascribingmeaning to novel stimuli. Curr. Biol. 18, R410–R412. doi: 10.1016/j.cub.2008.02.073

Mahon, B., and Caramazza, A. (2009). Concepts and categories: a cognitiveneuropsychological perspective. Annu. Rev. Psychol. 60, 27–51. doi: 10.1146/annurev.psych.60.110707.163532

Marr, D. (1982). Vision: A Computation Investigation into the Human Representationand Processing of Visual Information. San Francisco, CA: Freeman.

Mechelli, A., Sartori, G., Orlandi, P., and Price, C. J. (2006). Semantic relevanceexplains category effects in medial fusiform gyri. Neuroimage 30, 992–1002. doi:10.1016/j.neuroimage.2005.10.017

Op de Beeck, H. P., Baker, C. I., DiCarlo, J. J., and Kanwisher, N. (2006).Discrimination training alters object representations in human extrastri-ate cortex. J. Neurosci. 26, 13025–13036. doi: 10.1523/JNEUROSCI.2481-06.2006

Op de Beeck, H. P., Torfs, K., and Wagemans, J. (2008). Perceived shape sim-ilarity among unfamiliar objects and the organization of the human objectvision pathway. J. Neurosci. 28, 10111–10123. doi: 10.1523/JNEUROSCI.2511-08.2008

Perrett, D. I., and Oram, M. W. (1993). Neurophysiology of shape processing. ImageVis. Comput. 11, 317–333. doi: 10.1016/0262-8856(93)90011-5

Pylyshyn, Z. (1999). Is vision continuous with cognition? The case for cogni-tive impentrability of visual perception. Behav. Brain Sci. 22, 341–423. doi:10.1017/S0140525X99002022

Reber, R., Schwarz, N., and Winkielman, P. (2004). Processing fluency and aestheticpleasure: Is beauty in the perceiver’s processing experience? Pers. Soc. Psychol. 8,364–382. doi: 10.1207/s15327957pspr0804_3

Rhodes, G., Proffitt, F., Grady, J. M., and Sumich, A. (1998). Facial sym-metry and the perception of beauty. Psychon. Bull. Rev. 5, 659–669. doi:10.3758/BF03208842

Riesenhuber, M., and Poggio, T. (1999). Hierarchical models of object recognitionin cortex. Nat. Neurosci. 2, 1019–1025. doi: 10.1038/14819

Rosch, E., Mervis, C. B., Gray, W. D., Johnson, D. M., and Boyes-Braem, P. (1976).Basic objects in natural categories. Cogn. Psychol. 8, 382–439. doi: 10.1016/0010-0285(76)90013-X

Schwoebel, J., and Srinivas, K. (2000). Recognizing objects seen from novel view-points: effects of view similarity and time. J. Exp. Psychol. Learn. Mem. Cogn. 26,915–928. doi: 10.1037/0278-7393.26.4.915

Scott, L. S., Tanaka, J. W., Sheinberg, D. L., and Curran, T. (2006). A reevaluation ofthe electrophysiological correlates of expert object processing. J. Cogn. Neurosci.18, 1453–1465. doi: 10.1162/jocn.2006.18.9.1453

Scott, L. S., Tanaka, J. W., Sheinberg, D. L., and Curran, T. (2008). The roleof category learning in the acquisition and retention of perceptual exper-tise: a behavioral and neurophysiological study. Brain Res. 1210, 204–215. doi:10.1016/j.brainres.2008.02.054

Stroop, J. R. (1935). Studies of interference in serial verbal reactions. J. Exp. Psychol.18, 643–662. doi: 10.1037/h0054651

Tanaka, J. W. (2001). The entry point of face recognition: evidence for faceexpertise. J. Exp. Psychol. Gen. 130, 534–543. doi: 10.1037/0096-3445.130.3.534

Tanaka, J. W., Curran, T., and Sheinberg, D. (2005). The training and transfer ofreal-world perceptual expertise. Psychol. Sci. 16, 145–151. doi: 10.1111/j.0956-7976.2005.00795.x

Tanaka, J. W., and Taylor, M. (1991). Object categories and expertise: is the basiclevel in the eye of the beholder? Cogn. Psychol. 23, 457–482. doi: 10.1016/0010-0285(91)90016-H

Tarr, M. J., and Pinker, S. (1989). Mental rotation and orientation-dependencein shape recognition. Cogn. Psychol. 21, 233–282. doi: 10.1016/0010-0285(89)90009-1






Voss, J. L., Federmeier, K. D., and Paller, K. A. (2012). The potato chip reallydoes look like Elvis! Neural hallmarks of conceptual processing associated withfinding novel shapes subjectively meaningful. Cereb. Cortex 22, 2354–2364. doi:10.1093/cercor/bhr315

Wisniewski, E., and Medin, D. (1994). On the interaction of theory anddata in concept learning. Cogn. Sci. 18, 221–281. doi: 10.1207/s15516709cog1802_2

Witzel, C., Valkova, H., Hansen, T., and Gegenfurtner, K. (2011). Object knowledgemodulates colour appearance. i-Perception 2, 13–49. doi: 10.1068/i0396

Wong, A. C.-N., Palmeri, T. J., and Gauthier, I. (2009). Conditions for facelikeexpertise with objects: becoming a Ziggerin expert–but which type? Psychol. Sci.20, 1108–1117. doi: 10.1111/j.1467-9280.2009.02430.x

Wong, Y. K., Folstein, J. K., and Gauthier, I. (2011). Task-irrelevant perceptualexpertise. J. Vis. 11:3. doi: 10.1167/11.14.3

Wong, Y. K., Folstein, J. R., and Gauthier, I. (2012). The nature of experience deter-mines object representations in the visual system. J. Exp. Psychol. Gen. 141,682–698. doi: 10.1037/a0027822


Received: 26 March 2014; paper pending published: 27 April 2014; accepted: 06 July2014; published online: 29 July 2014.Citation: Cheung OS and Gauthier I (2014) Visual appearance interacts with con-ceptual knowledge in object recognition. Front. Psychol. 5:793. doi: 10.3389/fpsyg.2014.00793This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Cheung and Gauthier. This is an open-access article distributedunder the terms of the Creative Commons Attribution License (CC BY). The use, dis-tribution or reproduction in other forums is permitted, provided the original author(s)or licensor are credited and that the original publication in this journal is cited, inaccordance with accepted academic practice. No use, distribution or reproduction ispermitted which does not comply with these terms.












ORIGINAL RESEARCH ARTICLEpublished: 07 January 2015

doi: 10.3389/fpsyg.2014.01500

Segmentation of dance movement: effects of expertise,visual familiarity, motor experience and musicBettina E. Bläsing1,2,3*

1 Faculty of Psychology and Sport Science, Neurocognition and Action Research Group, Bielefeld University, Bielefeld, Germany2 Center of Excellence - Cognitive Interaction Technology, Bielefeld University, Bielefeld, Germany3 Research Institute for Cognition and Robotics (CoR-Lab), Bielefeld University, Bielefeld, Germany

Edited by:


Reviewed by:

Juliane J. Honisch, University ofBirmingham, UKMáximo Trench, Universidad delComahue, Argentina

*Correspondence:

Bettina E. Bläsing, Neurocognitionand Action Research Group,Bielefeld University,PO Box 100 131, 33501 Bielefeld,Germanye-mail: [email protected]

According to event segmentation theory, action perception depends on sensory cues andprior knowledge, and the segmentation of observed actions is crucial for understandingand memorizing these actions. While most activities in everyday life are characterized byexternal goals and interaction with objects or persons, this does not necessarily applyto dance-like actions. We investigated to what extent visual familiarity of the observedmovement and accompanying music influence the segmentation of a dance phrase indancers of different skill level and non-dancers. In Experiment 1, dancers and non-dancersrepeatedly watched a video clip showing a dancer performing a choreographed dancephrase and indicated segment boundaries by key press. Dancers generally defined lesssegment boundaries than non-dancers, specifically in the first trials in which visualfamiliarity with the phrase was low. Music increased the number of segment boundaries inthe non-dancers and decreased it in the dancers. The results suggest that dance expertisereduces the number of perceived segment boundaries in an observed dance phrase, andthat the ways visual familiarity and music affect movement segmentation are modulatedby dance expertise. In a second experiment, motor experience was added as factor, basedon empirical evidence suggesting that action perception is modified by visual and motorexpertise in different ways. In Experiment 2, the same task as in Experiment 1 wasperformed by dance amateurs, and was repeated by the same participants after they hadlearned to dance the presented dance phrase. Less segment boundaries were defined inthe middle trials after participants had learned to dance the phrase, and music reducedthe number of segment boundaries before learning. The results suggest that specificmotor experience of the observed movement influences its perception and anticipationand makes segmentation broader, but not to the same degree as dance expertise on aprofessional level.

Keywords: expertise, event segmentation, dance, movement learning, motor experience, music

INTRODUCTIONDespite its continuous nature, human motor action is func-tionally based on task- and event related perception. Researchsuggests that ongoing processing resources are devoted to thisperceptual process, and that the online perception of events deter-mines how episodes are understood and encoded in memory(Zacks and Tversky, 2001; Kurby and Zacks, 2008). According toEvent Segmentation Theory (Zacks et al., 2007), the perceptionof events depends on both sensory cues and knowledge structuresthat represent previously learned information about event partsand inferences about actors’ goals and plans. Related studies haverevealed that the segmentation of observed actions is crucial forthe understanding and memorizing of these actions (e.g., Swallowet al., 2009; Zacks et al., 2009; Sargent et al., 2013). Furthermore,the theory states that any observed activity is spontaneously seg-mented into events during perceptual processing, which enablesthe system to anticipate upcoming information and react appro-priately. As long as anticipation is successful, representations in

working memory (named “event models” in this context, seeZacks et al., 2007; Kurby and Zacks, 2008) are maintained ina stable state, guiding further prediction and saving process-ing costs. When the frequency of anticipation errors increasesas prediction becomes more difficult, event models are updatedbased on incoming information; these instances of increasinginsecurity are subjectively experienced as boundaries betweenevents. Perception of common goal-directed activities has beenfound to be hierarchical, with coarse-grained and fine-grainedsegmentation layers, corresponding to the hierarchical structureof action organization with goals and sub-goals. Furthermore,perception has been described as cyclical, with ongoing compar-ison of predictions to the perceived feeding back into processing.Event segmentation thereby results from the ongoing anticipa-tion of what will happen next, which serves action understanding,prediction and learning.

Studies using event segmentation paradigms have shown thatsegmentation characteristics can be related to the understanding

www.frontiersin.org January 2015 | Volume 5 | Article 1500 | 64












Bläsing Segmentation of dance movements

and memory of the observed actions. In these studies, actionsfrom everyday life, such as assembling objects, setting a table orfolding laundry, were presented to participants with no speci-fied expertise (e.g., Zacks et al., 2009; Sargent et al., 2013). Thepresented actions typically involve the manipulation of objectsand/or interactions between people, and are defined by clearaction goals and a clear semantic context. In the context of danceor sports, the same characteristics do not necessarily apply. Eventhough many skilled actions in a sports context are object- andperson-related and have clearly defined goals (e.g., passing the ballto a team member), there are also many other examples of move-ments that do not share these features. As such motor actionsoccur particularly in dance, the term “dance-like actions” hasbeen used to describe motor actions that lack common featuresascribed to actions from everyday contexts, such as interactionswith objects and persons and obvious external action goals.

It has been stated that the goal of such dance-like actions is “themovement itself,” which certainly is often the case in a dance con-text. Schachner and Carey (2013) showed that observers tended tointerpret actions as being intentionally movement-related if theywere not able to infer external goals from observing the action,or if the action seemed to be inefficient or inappropriate withregards to any recognizable external goal. The authors state thata dance-like action is also (in the eye of the observer) primar-ily characterized by its goal, which is movement-based, whereasother “rational” actions have external goals. This is particularlytrue for dance movements, the goal of which is commonly notonly movement-based but also related to communicating to part-ners or an audience via the body. It can therefore be assumedthat segmentation of dance-like actions or dance movementsfollows different “rules” and cognitive strategies compared tosegmentation of typical everyday activities with external goals.

Specifically in modern and contemporary dance, movementperformance often requires a fluent quality that does not affordobvious partitioning or segmenting. The ability to perform longmovement phrases with this obvious fluency is an important skillin these dance disciplines. This fluent quality of the movementcan be supported and enhanced by the accompanying music orsound. Choreographers might choose music that does not havea clear beat or rhythm but that rather provides an associatedsound layer, allowing the dancer and the spectator to integratemore freedom in integrating sound and movement. This meansthat the dance movement, when accompanied by music at all,does not necessarily follow a musical beat or rhythm, and mighteven contravene the music in order to create a more excitingimpression for the audience. The interrelation between the dancemovement and the accompanying music deliberately influencesthe spectator’s perception and should therefore be taken intoaccount when investigating the segmentation of dance move-ment; in this respect, dance differs from dance-like actions thatare not commonly associated with music.

Numerous studies have provided evidence that the percep-tion of skilled actions is modulated by expertise (see Cheung andBar, 2012) and is specifically facilitated by motor experience ofthe observed action type (e.g., Abernethy and Zawi, 2007; Agliotiet al., 2008; Güldenpenning et al., 2011; Steggemann et al., 2011).Even though empirical approaches to expertise often differentiate

between perception, cognition (e.g., decision making) and action(motor control), this distinction can hardly be maintained inthe context of athletes’ practice-dependent task-specific skills (seeYarrow et al., 2009). Evidence for the interdependency of per-ception, action and cognition in movement expertise has beenfound in many studies with athletes and other movement experts(e.g., Aglioti et al., 2008 see also Yarrow et al., 2009 for review).Dance expertise in particular has been shown to comprise a mul-titude of perceptual-motor and cognitive skills, including motorcontrol, timing, learning, memorizing, imagery, entrainment, aswell as multimodal communication and artistic expression (seeSevdalis and Keller, 2011; Bläsing et al., 2012; Waterhouse et al.,2014). Studies with expert dancers have shown that movement-related memory is more functionally structured in dancers than innon-dancers (Bläsing et al., 2009; Bläsing and Schack, 2012), andthat dancers show shorter fixation times while watching dancemovements than non-dancers, which points toward perceptionfacilitation (Stevens et al., 2010). Furthermore, dance provides ahighly adequate framework for studying expertise effects relatedto action-perception coupling, because dance, more than mosttypes of sports, is performed with the primary goal of beingobserved by an audience. In dancers, increased activity has beenfound in specific brain areas commonly referred to as actionobservation network (AON) while watching familiar dance move-ments (Calvo-Merino et al., 2005). This network of brain regions(comprising the ventral and dorsal premotor cortices and partsof the parietal cortex, including the inferior parietal lobe, thesuperior parietal lobe, and the superior parietal sulcus, as wellas the superior temporal sulcus) is typically involved in the exe-cution, observation and imagery of actions. Studies showed thatthe activation of these regions is modulated differently by visualand motor expertise (Calvo-Merino et al., 2006). Dancers showedincreased activity in areas belonging to the AON while watchingmovements from their own dance discipline compared to simi-lar movements from other dance disciplines (Calvo-Merino et al.,2005; Cross et al., 2006). Activation of AON regions was furtherincreased when dancers watched movements they had previouslyperformed themselves, compared to movements they had fre-quently watched but not physically performed (Calvo-Merinoet al., 2006). Learning to dance a specific movement phrase affectsAON activation while watching the same phrase already early dur-ing the learning process (Cross et al., 2006). Different types oflearning have been found to activate the AON in specific ways,with the right ventral premotor cortex responding specifically tothe experience of having performed an observed movement, andthe bilateral superior temporal cortex responding to the pres-ence of a human model (Cross et al., 2009). These findings reflectthat dance expertise affects both the production and the percep-tion of dance-like movements. Dance expertise should thereforenot only enable dancers to perform movement phrases fluently,but should also influence their perception of observed movementmaterial in favor of fluency and greater over-all connectedness.Only few studies have investigated the segmentation of dance-like actions (e.g., Pollick et al., 2012; Noble et al., 2014), and sofar none has focused on effects of dance expertise on segmenta-tion. Evidence from preliminary studies suggests that observers’dance expertise affects the segmentation of dance-like actions,

Frontiers in Psychology | Cognition January 2015 | Volume 5 | Article 1500 | 65





but not of other actions that have an obvious external goal(Bläsing et al., 2010).

The aim of the present study was to investigate how differentfactors, namely dance expertise, visual familiarity (via repeatedpresentations), motor experience (via learning to dance the pre-sented phrase) and music would influence the segmentation ofobserved dance movement. Specifically, we presented a chore-ographed contemporary dance phrase of fluent character that didnot contain interactions with objects or persons, communicativesignals or semantic content. The dance phrase was choreographedon the basis of modern/contemporary dance technique, and wasinitially novel to all participants. This means that participantswho regularly trained modern contemporary dance were likelyto be familiar with the type of movement in general, but notwith the presented movement material as such (note that mod-ern/contemporary dance choreography commonly involves theexploration and creation of new movement material rather thanre-combination and variation of defined partial movements, asthis is often the case in classical dance). During the experimentalprocedure, the participants watched the sequence repeatedly andbecame thereby increasingly familiar with it. Their visual familiar-ity, in terms of knowing the exact dance phrase (rather than sim-ilar movement material from the same disciplinary background)was addressed here as a factor potentially affecting segmentation.It was expected that segmentation would become less variablewith increasing visual familiarity over consecutive trials.

The issue of dance expertise was addressed in the current studyby comparing groups of participants differing in their specific skilllevel in dance. In Experiment 1, professional dancers who hadundergone professional dance training for many years and werecurrently all members of a professional dance theater companyor free-lancing professional dancers performing with differentcompanies as well as teaching dance on different levels (theseparticipants are in the following referred to as “dancers”) werecompared to sports students (in the following referred to as “non-dancers”) who had no particular experience in dance trainingapart from few very basic mandatory courses in their study pro-gram. It has to be pointed out that the non-dancer participantswere “novices” only with respect to dance, but not to movementskills in general; most of them performed their preferred sports onan advanced to high level. For the purpose of the study, this groupwas preferred to a group of participants without any movementexpertise (i.e., persons who did not perform sports or physicalexercise on a regular basis) because of the specific segmentationtask. Expertise has been shown to be task-specific and does notgeneralize well over domains (e.g., Ericsson and Charness, 1994).It was assumed that athletes without dance expertise would showsimilar responses to observed human body movement in generalcompared to dancers, including corresponding levels of motoractivation and simulation, and that any differences in the resultscould be related to expertise in dance rather than a high levelof physical training and motor skill in general. It was expectedthat dancers’ segmentation behavior would differ from that ofthe non-dancers, with dancers defining less segment boundariesbased on their training-based ability to anticipate dance move-ment more successfully despite its novelty, and their preferencefor viewing the observed dance movement as more connected.

As a third group, dance amateurs who trained mod-ern/contemporary dance on intermediate level participated inExperiment 2 of the presented study. These participants (referredto as “amateurs” in the following) were chosen for two reasons.First, they represented a viable intermediate step between thenon-dancers and the dancers, offering the opportunity to monitorexpertise effects on different levels. Second, the amateur par-ticipants all belonged to the same dance class that was trainedby the choreographer of the stimulus dance phrase. Crucially,this class was taught the dance phrase as part of their trainingschedule, which provided the opportunity to add the aspect oflearning to that of expertise and relate the two aspects to eachother. In Experiment 2, the participants thereby gained specificmotor experience of the presented movement material (referredto as “motor experience” in the following, applied as factor inExperiment 2). The term “motor experience” is in this case relatedto the experience of having danced the exact phrase presented asstimulus, not more generally to experience with similar move-ment material from the same disciplinary background. It wasexpected that specific motor experience would increase the par-ticipants’ expertise for the dance phrase and thereby make theirsegmentation behavior more “expert-like,” potentially even morethan the dancers’ in Experiment 1 who had greater dance exper-tise in general but no motor experience of the presented dancephrase.

As a fourth factor, the presence of music was added. Themusic chosen by the choreographer to accompany the dancephrase did not have a clear metric rhythm but rather con-sisted of an underlying sound layer of chords with slowlyincreasing and decreasing pitch and volume. It was hypothe-sized that the added music, because of its specific character,would influence the segmentation of the movement by reduc-ing the number of segment boundaries, thereby binding move-ments together and reducing the over-all number of segmentboundaries.

EXPERIMENT 1: SEGMENTATION OF A DANCE PHRASE BYDANCERS AND NON-DANCERS: EFFECTS OF VISUALFAMILIARITY AND MUSICIn Experiment 1, professional dancers and sport students withoutdance expertise repeatedly watched a video clip showing a dancerperforming a phrase from a contemporary dance choreography.Each participant watched the sequence 20 times on a computerscreen, 15 times without music followed by 5 times with music,and indicated segment boundaries by key press. This experimentwas conducted in order to gain information about the effects ofdance expertise, visual familiarity and music on the segmentationof dance movement.

METHODParticipantsTwenty-two participants voluntarily took part in Experiment 1without any exchange for course credit or money. Twelve studentsof sport science (six females, one left-handed; age 25.91 ± 3.29years, range 22–30 years) without any particular dance trainingexperience (except for basic courses as part of their study pro-gram) were assigned to the non-dancers’ group. All non-dancers





were physically active; their most regularly performed sportsincluded soccer, handball, rugby, and fitness training.

Ten professional dancers (six females, two left-handed; age30.1 ± 6.59, range 23–40 years) participated as experts; allwere trained in classical, modern and contemporary danceon professional level and were currently active as companydancers. Six of the dancers were current members of TanztheaterBielefeld; four of the dancers were freelancing dancers and danceteachers.

All participants reported having normal or corrected-to-normal vision, and were naive with regard to the purpose of theexperiment. All participants provided written informed consentbefore testing started. The experiment was performed in accor-dance with the ethical standards of the sixth revision (WMA,2008) of the 1964 Declaration of Helsinki.

Apparatus and StimuliThe stimulus material consisted of a video clip (92 s, 2.290frames, 25 Hz, recorded with a Sony camcorder) showing a dancephrase created and performed by dancer and choreographer IlonaPászthy. The dance phrase was choreographed on the basis ofmodern/ contemporary dance technique, and was novel to all par-ticipants. For stimulus presentation and data collection, Interact®(Mangold) software running on a Notebook (Acer) with a 15inch VGA-Display (vertical retraces 60 Hz) was used. The soft-ware recorded key presses during the presentation of the videoclip, linked them to the adequate runtime and frame number andprovided a protocol of these data.

Design and ProcedureThe data collection took place in a quiet lab or office room or ina free rehearsal space at the theater. Each participant was testedindividually. During the experiment, the participant sat in frontof the notebook computer and watched the presented video clip.The following instructions were given verbally by the experi-menter: “You will now see a video clip of a dancer dancing apart of a dance piece. The clip will be repeated 20 times. Whilewatching, please keep your finger on the space bar and press thespace bar each time a part of the dance phrase ends and a newone begins. Apply your own criteria; you do not need to mark thesame moments in each repetition.” This instruction was phrasedin a similar way as instructions in previous segmentation studies(e.g., “to press a button... whenever... one natural and meaning-ful unit of activity ended and another began,” Zacks et al., 2009).No instruction was given regarding the resolution of segmenting(fine or coarse), as had been done in other segmentation studies(e.g., Swallow et al., 2009; Zacks et al., 2009). The sequence waspresented 20 times, the first 15 trials without sound, followed byfive trials accompanied by the music that had been chosen by thechoreographer.

After completing all 20 trials, the participant was verballyasked two questions by the experimenter:

1. Which criteria or strategies did you use for segmenting thedance phrase?

2. Did the music in the last five trials affect your decisions?

The answers were written down by the experimenter in the formof key notes. This explorative interview was not carried outaccording to any established qualitative method, but was added tothe data collection only to gain an impression of the participants’use of criteria and strategies. It was not expected that participantswould be able to give a complete and objective account of theirsegmenting behavior, but the experimenter was rather interestedin the criteria and strategies the participants applied explicitlyor even deliberately. The complete experimental session for eachparticipant lasted 60–90 min.

Data analysesFor every participant, the number of segment boundaries wasrecorded for each of the 20 trials. Mean group results of dancersand non-dancers were calculated for each trial number (1–20)separately, for all trials together, and for four groups of trials(trials 1–5: early trials; these trials were regarded as familiar-ization phase during which visual familiarity with the dancephrase was still low; trials 6–10: middle trials, with increas-ing visual familiarity; trials 11–15: late trials; for these tri-als, visual familiarity with the dance phrase was regarded ashigh; and trials 16–20: music trials, presented with sound).Non-parametric tests were applied to compare dancers and non-dancers regarding their defined numbers of segment bound-aries for each trial separately, for all trials, and for each groupof five trials (early trials, middle trials, late trials and musictrials). Within each group of participants, mean numbers ofsegment boundaries of the four trial groups (early, middle,late, and music) were compared to each other using non-parametric tests (Mann Whitney U-test, Wilcoxon signed-ranktest).

RESULTSSegment boundariesComparisons between dancers and non-dancers (Mann-WhitneyU-test) revealed that dancers generally defined less segmentboundaries than non-dancers for all trials together (z = −2.853,p = 0.005), for each individual trial (trials 1–5: p < 0.01; tri-als 6–13: p < 0.05; trials 14–20: p < 0.01), and for all groupsof trials (early trials: z = −3.269, p = 0.001; middle trials:z = −2.474, p = 0.013; late trials: z = −2.440, p = 0.015; musictrials: z = −2.969, p < 0.003). Comparisons between groups oftrials (Wilcoxon signed-rank test) in the non-dancers revealeddifferences between middle trials and music trials (z = −2.296,p = 0.022) and between late trials and music trials (z = −2.173,p = 0.030), with more segment boundaries occurring in themusic trials than in the other groups. In the dancers, less segmentboundaries were defined in the early trials than in the middletrials (z = −2.018, p = 0.044). In contrast to the non-dancers’results, less segment boundaries were defined in the music tri-als than in the late trials (z = −2.092, p = 0.036). Results forthe four groups of trials are displayed in Figure 1, results for allindividual trials are shown in Figure 3. The distribution of seg-ment boundaries (calculated as average over all trials for 92 binsof 1 s) as defined by the experimental groups is illustrated inFigure 4.






FIGURE 1 | Results of Experiment 1. Mean numbers of segmentboundaries defined for the four groups of trials [1: trials 1–5 (early); 2: trials6–10 (middle); 3: trials 10–15 (late); 4: trials 16–20 (music)]; blue columns:dancers, red columns: non-dancers; asterisks mark significant differences:∗p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001.

Post-hoc interviewsAfter finishing the experimental procedure, each participant wasasked two questions:

1. Which criteria or strategies did you use for segmenting thedance phrase?

2. Did the music in the last five trials affect your decisions?

The experimenter asked the participant verbally and wrote downthe answers in key points (note that this informal procedure didnot follow any established qualitative approach but only aimed atgaining additional information in an explorative way).

From the informal answers to Question 1, the most commoncriteria were extracted, and naming frequencies of these criteriawere counted. The most common criteria and their frequenciesof naming are displayed in Table 1. Remarks made by individ-ual participants in response to Questions 1 and 2 are listed inSupplementary Table 1.

DISCUSSIONIn Experiment 1, dancers and non-dancers segmented a dancephrase repeatedly presented in a video clip by key press.Segmentation grain (i.e., numbers of segment boundaries) wasexpected to be influenced by expertise (comparison between thetwo groups), by visual familiarity of the movement phrase (com-parison between early, middle and late trials), and by music(comparing the last group of trials presented with music to theprevious groups of trials).

The results showed that in all trials, in each individual trialand in the four groups of trials, dancers generally defined lesssegment boundaries than non-dancers. The effect of expertiseon movement segmentation was thereby very clearly reflectedby the results, with dancers defining less segment boundariesand thereby segmenting the whole movement phrase into fewer

Table 1 | Segmentation criteria named by the three groups of

participants in the post-hoc interviews (numbers indicate absolute

frequencies of naming in both experiments).

Segmentation criteria Non-dancers Amateurs Dancers

Change of movement type 7 8 4

Change of height level 3 3 6

Stops, pauses 5 2 3

Change of direction in space 4 2 3

Change of main active body part 4 1 3

Change of tempo, dynamics – 8 5

Feeling, imagery – 5 4

Movement impulse, accents – 3 2

Cues for learning or teaching – 5 –

Change of energy or force – – 4

and longer sections than non-dancers. This finding is supportedby the comment of one dancer, who reported perceiving theentire phrase as a whole, “in a flow,” therefore segmenting didnot feel natural. Perceiving a longer dance phrase as a wholedespite the occurrence of various movement characteristics thatcould be used (and were typically named) as segmentation cri-teria is also in accordance with the claim often made in modernand contemporary dance to dance longer phrases fluently with-out obvious breaks or partitions, without “losing the energy.”The finding that this principle commonly applied to the dancers’action performance is transferred to perception when observinga dance phrase accords with the principle of perceptual resonance(Schütz-Bosbach and Prinz, 2007) described in various areas ofexpertise (e.g., Kiesel et al., 2009; Güldenpenning et al., 2011;Steggemann et al., 2011). Dancers defined less segment bound-aries in early trials than in the middle and late trials, whereasno difference between early, middle and late trials was found inthe non-dancers. An effect of visual familiarity was thereby onlyfound in the dancers, but not in the non-dancers.

Interestingly, music affected segmentation differently indancers and non-dancers: In the music trials, dancers definedless segment boundaries than in late trials, whereas non-dancersdefined more segment boundaries in music trials than in mid-dle and late trials. Apparently, music had a binding effect on theperceived movement in the dancers’ group. (Comments givenby individual dancers in response to Question 2 supported thisinterpretation: music was experienced as binding the movementtogether, slowing down the movement, adding a harmonic feel-ing). In the non-dancers, in contrast, music seemed to confuseand thereby cause more segment boundaries to occur, possiblybased on the perceived lack of segmentation cues in the music thatmight have interfered with previously defined movement cues.

Expertise in sport or dance typically involves visual as wellas motor experience of specific actions, and differences foundbetween experts and novices can be based on any of the two, orboth. To gain further understanding of expertise effects in actionperception, it is necessary to differentiate visual and motor exper-tise either by studying observation experts (e.g., Calvo-Merinoet al., 2006; Aglioti et al., 2008) or by applying a learning interven-tion (e.g., Cross et al., 2006, 2009). In order to gain information





about potential effects of motor experience on segmentation,a second experiment was conducted with dance amateurs whosolved the same experimental task as applied in Experiment 1before and after learning the presented dance phrase.

EXPERIMENT 2: SEGMENTATION OF A DANCE PHRASEBEFORE AND AFTER LEARNING: EFFECTS OF MOTOREXPERIENCE, VISUAL FAMILIARITY AND MUSICIn Experiment 2, the same segmentation task as in Experiment1 was applied to dance amateurs who regularly trained moderndance in the same group. After learning the phrase in their train-ing as part of a performance program, all participants repeatedthe experimental task. The main goal of the experiment, apartfrom adding a third (intermediate) group of participants, was togain information regarding the effect of specific motor expertiseon segmenting a dance phrase.

METHODParticipantsEight participants (all female, one left-handed; age 18.5 ± 6.55years, range 14–30 years) voluntarily took part in Experiment 2without any exchange for course credit or money. All participantstrained regularly in classical and contemporary dance on averageto advanced amateur level (years of training in classical dance:9.13 ± 1.45 years, range: 8–12 years; years of training in mod-ern dance: 3.38 ± 1.51 years, range: 1–6 years; dance training:3.38 ± 2.20 h per week, range: 2–6 h) and were currently mem-bers of the same modern dance class (Theater Bielefeld balletschool). All eight participants named classical and modern danceas their primary types of training, single participants also trainedin one of the following disciplines: capoeira, hip-hop, karate andacrobatics. All participants reported having normal or corrected-to-normal vision, and were naive with regard to the purpose ofthe experiment. All participants provided written informed con-sent before testing started. The experiment was performed inaccordance with the ethical standards of the sixth revision (WMA,2008) of the 1964 Declaration of Helsinki.

Apparatus and StimuliThe same stimulus material and experimental set-up was used asin Experiment 1.

Design and ProcedureTwo data collections were applied, one before and one afterthe participants learned the dance phrase in their training.Data collections took place in a quiet lab or office room orin a free dress room at the ballet school. Each participant wastested individually, the experimental procedure was exactly thesame as in Experiment 1. The single experimental session lasted60–90 min.

After all participants had completed the experiment once (datacollection 1, pre-learning), they learned the presented dancephrase as part of their regular training. After approximately 6weeks in which the dance phrase had been trained regularly, theexperiment was repeated in exactly the same way as before (datacollection 2, post-learning). Crucially, at the time of data collec-tion 1, participants were neither informed that they would learnthe dance phrase nor that they would be asked to participate in a

second data collection, and participants were not informed aboutdata collection 2 when learning the dance phrase. The data collec-tions were separated by a time interval of approximately 6 weeksduring which the dance phrase was learned and trained as part ofa choreography for later stage performance.

Data AnalysesMean numbers of segment boundaries were analyzed in the sameway as for Experiment 1. Non-parametric tests were applied tocompare the two experimental conditions, pre- and post-learning(i.e., without and with motor experience of dancing the phrase,respectively), regarding the defined numbers of segment bound-aries for each trial separately, for all trials, and for each group offive trials (early trials, middle trials, late trials, and music trials).For each data collection (pre- and post-learning), mean numbersof segment boundaries of the four trial groups (early, middle, late,and music) were compared to each other using non-parametrictests (Wilcoxon signed-rank test).

RESULTSSegment boundariesComparisons of trial groups between the pre- and post-learningconditions (Wilcoxon signed-rank test) revealed a difference inthe middle trials (z = −2.240, p = 0.025), in which less segmentboundaries were defined in the post-learning condition thanin the pre-learning condition. Comparisons of individual trialsrevealed differences in trials 8, 9, 10, 12, and 15 (all p < 0.05). Nodifference between pre- and post-learning was found, however,when comparing segment boundaries over all trials. Comparisonsbetween groups of trials within each condition revealed thatonly in the pre-learning condition, less segment boundaries weredefined in the music trials than in the late trials (z = −2.371, p =0.018), whereas trial groups did not differ in the post-learningcondition. Results for the four groups of trials are displayed inFigure 2, results for all individual trials (Experiments 1 and 2) areshown in Figure 3. The distribution of segment boundaries (cal-culated as average over all trials for 92 bins of 1 s) as defined by theamateurs before and after learning the dance phrase is illustratedin Figure 4.

Comparing the group of amateurs to the groups of dancersand non-dancers from Experiment 1 showed differences betweenthe non-dancers and the amateurs in the pre-learning con-dition for the early trials (z = −2.013, p = 0.044) and themusic trials (z = −2.394, p = 0.017), and differences betweenthe non-dancers and the amateurs in the post-learning condi-tion in all groups of trials (early: z = −2.355, p = 0.019; mid-dle: z = −2.431, p = 0.015, late: z = −2.546, p = 0.011, music:z = −3.009, p = 0.003), as well as over all trails (z = −2.508,p = 0.012). No differences were found between the dancers’ andthe amateurs’ results. Results for all individual trials are displayedin Figure 3.

Post-hoc interviewsAs in Experiment 1, each participant was verbally asked twoexplorative questions after each data collection (again, no estab-lished qualitative approach was applied but the two questionswere asked informally and key points of the answers were written






FIGURE 2 | Results of Experiment 2. Mean numbers of segmentboundaries defined for the four groups of trials (1, early trials; 2, middletrials; 3, late trials; 4, music trials); blue columns: before learning, redcolumns: after learning; asterisks mark significant differences: ∗p < 0.05.

FIGURE 3 | Results of Experiment 1 (full lines) and Experiment 2

(dashed lines) combined, mean numbers of segment boundaries

defined for individual trials. Red, non-dancers; green, dancers; light blue,amateurs before learning; dark blue, amateurs after learning.

down by the experimenter). The most common criteria and theirfrequencies of naming (in response to Question 1: “Which crite-ria or strategies did you use for segmenting the dance phrase?”)are displayed in Table 1. Remarks made by individual participantsin response to Questions 1 and 2 (“Did the music in the last fivetrials affect your decisions?”) are listed in Supplementary Table 1.

DISCUSSIONIn Experiment 2, dance amateurs segmented a dance phraserepeatedly presented in a video clip by key press. The experi-ment was repeated after the participants had learned to dancethe presented dance phrase. Segmentation grain (i.e., numbersof segment boundaries) was expected to be influenced by motorexperience (comparison between pre- and post-learning), byvisual familiarity of the movement phrase (comparison betweenearly, middle and late trials), and by music (comparing the last

group of trials presented with music to the previous groups oftrials).

Results showed that less segment boundaries were defined inthe post-learning condition compared to the pre-learning condi-tion in the middle and late trials. No difference between pre- andpost-learning was found in the early trials and in trials with music,and when comparing mean numbers of segment boundaries overall trials. Consequently, the motor experience of dancing the pre-sented movement phrase was found to affect segmentation grainslightly, with less segment boundaries being defined by the par-ticipants after they had learned to dance the phrase, however,this difference only reached significance in the middle and latetrials. This finding was reflected by the participants’ comments:segments were perceived as longer, “larger shapes” were recog-nized, longer segments were “more fun.” Participants’ commentsalso reflected that watching the dance phrase was experienced asmore embodied and more competent after learning.

Music was found to affect segmentation in the pre-learningcondition, but not in the post-learning condition. Before learn-ing the dance phrase, less segment boundaries were defined inthe music trials than in the late trials, showing an effect of musicon segmentation comparable to the one observed in the dancers.It can be assumed that the effect was caused by a binding effectof music (as assumed in the dancers). The finding that the dif-ference between late trials and music trials was not significantanymore after learning the movement phrase might be explainedby the fact that the participants danced the phrase in their trainingin combination with the same music as used in the experiment,therefore music and movement might have been coupled duringthe learning and training process. When watching the movementwithout the music in the post-learning condition, participantsdid not really experience the movement “without the music,” asmusic and movement had become parts of the same integratedrepresentation in their long-term memory (see Land et al., 2013),and segmentation (even when no music was played) related notonly to the presented movement, but to the representation of“movement-with-music.”

In contrast to Experiment 1, no differences were foundbetween early, middle and late trials within each condition, show-ing that no effect of visual familiarity was found in the amateurs(or that the potential effect of visual familiarity was too weakto produce significant results). The finding that visual familiar-ity did not significantly affect segmentation is contrasted by theimpression of several participants that, with repeated observa-tions of the dance phrase in consecutive trials, the dance phrasewas more strongly perceived as a whole (the movement was“growing together”). Similar to the dancers, several amateurs hadexpressed before learning that they found it difficult to segmentthe phrase because of the perceived fluency and connectedness ofthe movement.

Comparing the results of the amateurs’ group in Experiment 2to the results of the two groups of participants from Experiment1 revealed that the amateurs did not differ from the dancers,whereas differences found between the amateurs and the non-dancers were found and increased from the pre-learning tothe post-learning condition. These findings indicate that theamateurs might have become more “expertly” in perception and





FIGURE 4 | Results of Experiments 1 and 2. Distribution of segmentboundaries over the phrase: average number of segment boundaries for eachbin (1 s), calculated over all trials; panels from top to bottom: dancers,non-dancers, amateurs before learning the phrase, amateurs after learningthe phrase. The most frequently set segmentation boundaries marked the

following instances in the dance phrase: 13 change of speed and directionduring a floor sequence; 29 slowing down and raising to the knees from floorsequence; 53 slowing down floor movement, raising into shoulder stand; 76transition from walking to dynamic swinging movement initiated by animpulse of the right shoulder; 86 going down to the floor from standing.

segmentation by learning to dance the presented movement,which corroborates findings on effects of motor expertise onaction perception (e.g., Aglioti et al., 2008; Güldenpenning et al.,2011).

GENERAL DISCUSSIONExpertise in dance and various sports disciplines has been foundto modulate the perception of actions on different levels, specif-ically of those actions belonging to the specific area of expertise(Calvo-Merino et al., 2005, 2006; Aglioti et al., 2008; Cheung andBar, 2012). Evidence exists that these effects specifically relate tomotor experience and learning of the observed actions (Crosset al., 2006, 2009). Furthermore, Event Segmentation Theory(Zacks et al., 2007) and related empirical studies (e.g., Zackset al., 2009; Sargent et al., 2013) provide evidence for the assump-tion that the segmentation of observed actions is influenced byrelevant visual and motor expertise. In the present study, thisassumption was tested using segmentation of a dance phrase asexperimental task performed by three groups of participants dif-fering in dance expertise. Results of two consecutive experimentsrevealed broader segmentation applied by professional dancers,

but also by dance amateurs, compared to non-dancers. The effectwas increased in the amateurs after learning the presented dancephrase, pointing toward a specific effect of motor experience. Ithas to be emphasized that in this study participants were notinstructed to segment with fine or coarse segmentation grain,whereas this was commonly done in studies on event segmen-tation. When participants were instructed to segment observedactions into coarse units, this typically resulted in segment lengthsof 30–60 s, whereas the instruction to apply fine-grained segmen-tation resulted in units of 10–20 s. Zacks et al. (2009) showedthat fine-grained segmentation generally correlated with move-ment parameters whereas coarse-grained segmentation rathercorrelated with external goals and context information. In thecurrent study, average segment lengths defined by the partici-pants ranged between 6 and 13 s. As the context of the presentedmovement in this study was dance and the dancer’s intentionwas clearly movement-related, it can be assumed that segmenta-tion was predominantly fine-grained, and the results support thisassumption.

Visual familiarity was induced in the current study by repeatedpresentation of the same stimulus dance phrase. This approach






clearly differs from the approaches taken in other segmentationstudies in which observes typically watched the same action twice,in part with different instructions (e.g., coarse vs. fine segmenta-tion, Zacks et al., 2009; Sargent et al., 2013; or just watching vs.segmenting, Noble et al., 2014). Under natural conditions, actionsare not repeated in exactly the same way, and segmentation occursspontaneously as part of perceptual processing. The same com-monly applies to watching dance: most audience members watchthe performance of a dance piece once, from a naive perspec-tive. “Expert observers” like dancers, choreographers, teachers,and dance enthusiasts watch the same movement phrases repeat-edly, however, performances naturally differ. Watching differentversions of the same movement material performed with nat-ural variation may increase the observers’ sensitivity for subtledifferences. Watching the identical performance 10 or 20 times(e.g., from a video clip) certainly increases the observer’s sensi-tivity for details, but might also become boring. In the currentstudy, the phrase was presented repeatedly to create high visualfamiliarity with the previously unknown phrase, thereby facili-tating prediction and increasing anticipation success. FollowingEvent Segmentation Theory (Zacks et al., 2007), this should resultin a decrease of the number of segment boundaries. Resultsof this study confirm this assumption, and individual partici-pants’ comments reflect this (see Supplementary Table 1). Otherparticipants’ comments suggest that certain segment boundariesbecame fixed over trials, as participants felt more confident tobe “right” in their decisions. This aspect of experiencing com-petence might play a crucial role for the esthetic evaluation ofdance, which has been related to prediction success and momentsof surprise (e.g., Hagendoorn, 2004). Studying segmentation inrelation to the novelty and esthetic evaluation of dance movementwould be highly relevant in this context, addressing for instanceobservers’ subjective experiences of boredom and competence. Inthe present study, several participants reported that they delib-erately varied their strategies, playfully trying out different waysof segmenting (see Supplementary Table 1). This creative atti-tude toward the experimental task might have been an attempt tocounteract boredom, but might also have been elicited by the typeof stimulus (dance) and the general appreciation of watching it.

Music added during the last five trials had contradictory effectsin the different experimental groups. While music decreasedthe number of segment boundaries in the dancers and in theamateurs (before learning), it increased the number of segmentboundaries in the non-dancers. This finding has been interpretedin terms of a binding effect in the dancers and amateurs andirritation in the non-dancers.

Empirical evidence exists that event segmentation not onlyoccurs while observing actions, but also while listening to music(Sridharan et al., 2007). Based on the proposed multimodalityof event models (see Zacks et al., 2007), it can be assumed thatmusic modifies the perception (and performance) of the dancemovement it accompanies, by potentially increasing the experi-ence of uncertainty in movement prediction at musical transitionpoints (Sridharan et al., 2007) and decreasing it within musi-cal phrases. The effect of this interrelation between perceivedmovement and sound on segmentation is likely to depend on thespecific characteristics of both and their temporal relation to each

other. As listening to music alone has been shown to be sensitiveto event segmentation (Sridharan et al., 2007), it can be assumedthat music would affect the segmentation of movement in differ-ent ways, depending on its characteristics (i.e., its metrics, pitch,rhythm, pulse, etc.). In the current study, the music that accompa-nied the movement did not have any metric rhythm or pulse, butconsisted of slowly rising and falling chords, and might thereforehave had a binding rather than a dividing influence. It can fur-thermore be hypothesized that segmentation of dance movementaccompanied by music would be sensitive to the way both areintegrated, and to what extent the dancer entrains with the music,reacts to musical cues or deliberately counterpoints them. Thisaspect and the more general question to what extent music influ-ences the segmentation of observed dance movement warrantsfurther study. It would be particularly interesting to systemati-cally combine movement and sound in different ways to shedlight on the roles of visual and auditive information and theirinterrelation for the perception and segmentation of dance-likeactions. Research responding to these questions would not onlybe of interest for scientists investigating action perception, butalso for dancers and choreographers interested in audience reac-tions. Promising manipulations could include the presentationof a movement phrase combined with different types of musicand sound, or with the same music or sound varying in temporalrelation (i.e., music systematically shifted relative to the move-ment). In such studies, the different combinations of music andmovement would have to be presented in a counterbalanced wayto control for order effects. In the current study, this was notthe case; music was added only to the last five trials to preventinterference with the factor visual familiarity. Confounding of thefactor music with visual familiarity can therefore not be excluded,which represents a clear limitation of the current study withregards to the influence of music on movement segmentation.Furthermore, as previously mentioned, the typical situation of adance spectator is to watch a dance performance once, withoutknowing it. Therefore, potential effects of music on the percep-tion of novel, unfamiliar and potentially surprising movementmaterial would be relevant to study in the context of dance.

The task of parsing movement phrases has also been appliedas artistic tool in choreography, as it is assumed to have thepotential to change the dancers’ perception of the movementand thereby their artistic expression. In a “parsing and view-ing” task performed by Wayne McGregor | Random Dance aspart of a choreographic process, dancers segmented movementphrases and subsequently commented their decisions, whichrevealed different cognitive frameworks underlying dance parsing(deLahunta and Barnard, 2005). To attend to the latter aspect inour study, the participants of the current study were asked infor-mally about their personal segmentation criteria and strategies.Participants of all groups named changes in movement character-istics (movement type, active body part, height level, direction,speed). Criteria related to learning the movement were onlynamed by the amateurs (e.g., “how the teacher would teach it”),whereas only the dancers named dynamic features (e.g., “whereforce would be needed”). It can be assumed that the latter crite-rion requires efficient on-line motor simulation of the observedmovement, which might be a skill that is specific to dancers on a





high level of expertise (see Bläsing and Schack, 2012). To addressthis issue, further research would be needed relating experts’ seg-mentation with movement analysis. Similar approaches have beenused by Zacks et al. (2009) for actions from a non-dance con-text and by Noble et al. (2014) for Indian dance with a narrativecharacter, but these studies did not include an expertise-relatedparadigm. It would be of particular interest to know to whatextent dancers, compared to non-dancers, are able to specify andpredict dynamical measures (i.e., forces) from motor simulationwhile watching dance movement, and how this influences the waythey segment a dance phrase.

An aspect of this study that has been rarely addressed isthe combination of comparing expertise with a learning task.Learning, however, was investigated only in the amateur groupwith intermediate skill level, but not on different levels of exper-tise, which could be a topic for further study. A related questionof interest that could not be addressed here is to which extent theway movement material is learned or taught affects the perceptionand segmentation of the same material later on. In other words,would the way the teacher has structured the dance phrase bereflected by the segment boundaries defined by the students afterlearning the phrase? In dance training, teachers commonly breakmovement phrases down into sub-phrases and partial movementsin order to facilitate learning, and students imitate and practicethese parts and subsequently combine them again to a wholephrase. This procedure naturally affords breakpoints in the men-tal representation of the phrase in the dancer’s memory that canbecome undesired breakpoints in the fluent quality of movementperformance. To cure this problem, different measures are takenduring further training of the phrase, such as varying the lengthsof partial phrases and paying special attention to the transitions.Studies using segmentation paradigms could help to shed light onthe effects and the efficiency of such practice.

Taken together, it can be concluded that segmentation ofdance movement is clearly influenced by expertise, with broadersegmentation grain being applied by professional and amateurdancers than non-dancers. Effects of visual familiarity and musicon movement segmentation seem to be modulated by expertise,and motor experience had a slight effect in the amateur dancers.These findings contribute to the literature on dance expertiseand segmentation of dance-like actions, and raise future researchquestions particularly addressing effects of the novelty or famil-iarity of the observed movement material, interrelations betweenmovement and music or sound, learning in different ways andon different levels of expertise, and the esthetic evaluation of theobserved dance movement.

ACKNOWLEDGMENTSThe author thanks the members of Tanztheater Bielefeld, theballet school of Theater Bielefeld and all participants for theirfriendly support. I am particularly grateful to Ilona Pászthy whochoreographed and danced the phrase presented in the experi-ments and trained the class of dance amateurs who participatedin Experiment 2. The author acknowledges support for the ArticleProcessing Charge by the Deutsche Forschungsgemeinschaft(DFG) and the Open Access Publication Funds of BielefeldUniversity Library.


01500/abstract

REFERENCESAbernethy, B., and Zawi, K. (2007). Pickup of essential kinematics underpins

expert perception of movement patterns. J. Mot. Behav. 39, 353–367. doi:10.3200/JMBR.39.5.353-368

Aglioti, S. M., Cesari, P., Romani, M., and Urgesi, C. (2008). Action anticipationand motor resonance in elite basketball players. Nat. Neurosci. 11, 1109–1116.doi: 10.1038/nn.2182

Bläsing, B., Calvo-Merino, B., Cross, E. S., Jola, C., Honisch, J., and Stevens, C.J. (2012). Neurocognitive control in dance perception and performance. ActaPsychol. 139, 300–308. doi: 10.1016/j.actpsy.2011.12.005

Bläsing, B., Kornfeld, C., and Schack, T. (2010). “Bewegungssegmentierung imtanz: effekt von expertise,” in Bewegung und Leistung: Sport, Gesundheit undAlter, 22 Edn., eds K. Mattes and B. Wollesen (Hamburg: Feldhaus Verlag; Band204), 22.

Bläsing, B., and Schack, T. (2012). Mental representation of spatial move-ment parameters in dance. Spat. Cogn. Comput. 12, 111–132. doi:10.1080/13875868.2011.626095

Bläsing, B., Tenenbaum, G., and Schack, T. (2009). The cognitive structure ofmovements in classical dance. Psychol. Sport and Exercise, 10, 350–360. doi:10.1016/j.psychsport.2008.10.001

Calvo-Merino, B., Glaser, D. E., Grèzes, J., Passingham, R. E., and Haggard, P.(2005). Action observation and acquired motor skills: an FMRI study withexpert dancers. Cereb. Cortex 15, 1243–1249. doi: 10.1093/cercor/bhi007

Calvo-Merino, B., Grèzes, J., Glaser, D. E., Passingham, R. E., and Haggard, P.(2006). Seeing or doing? influence of visual and motor familiarity in actionobservation. Curr. Biol. 16, 1905–1910. doi: 10.1016/j.cub.2006.07.065

Cheung, O. S., and Bar, M. (2012). Visual prediction and perceptual expertise. Int.J. Psychophysiol. 83, 156–163. doi: 10.1016/j.ijpsycho.2011.11.002

Cross, E. S., Hamilton, A. F., and Grafton, S. T. (2006). Building a motor simula-tion de novo: observation of dance by dancers. Neuroimage 31, 1257–1267. doi:10.1016/j.neuroimage.2006.01.033

Cross, E. S., Kraemer, D. J., Hamilton, A. F. D. C., Kelley, W. M., and Grafton,S. T. (2009). Sensitivity of the action observation network to physical andobservational learning. Cereb. Cortex 19, 315–326. doi: 10.1093/cercor/bhn083

deLahunta, S., and Barnard, P. (2005). “What’s in a Phrase?,” in Tanz im Kopf/ Dance and Cognition, eds J. Birringer and J. Fenger (Munster: LIT Verlag;Jahrbuch der Gesellschaft für Tanzforschung 15), 253–266.

Ericsson, K. A., and Charness, N. (1994). Expert performance: its structure andacquisition. Am. Psychol. 49, 725. doi: 10.1037/0003-066X.49.8.725

Güldenpenning, I., Koester, D., Kunde, W., Weigelt, M., and Schack, T. (2011).Motor expertise modulates the unconscious processing of human body pos-tures. Exp. Brain Res. 213, 383–391. doi: 10.1007/s00221-011-2788-7

Hagendoorn, I. (2004). Some speculative hypotheses about the nature and percep-tion of dance and choreography. J. Conscious. Stud. 11, 79–110.

Kiesel, A., Kunde, W., Pohl, C., Berner, M. P., and Hoffmann, J. (2009).Playing chess unconsciously. J. Exp. Psychol. Learn. Mem. Cogn. 35, 292. doi:10.1037/a0014499

Kurby, C. A., and Zacks, J. M. (2008). Segmentation in the perception and memoryof events. Trends Cogn. Sci. 12, 72–79. doi: 10.1016/j.tics.2007.11.004

Land, W. M., Volchenkov, D., Bläsing, B., and Schack, T. (2013). From actionrepresentation to action execution: exploring the links between cognitive andbiomechanical levels of motor control. Front. Comput. Neurosci. 7:127. doi:10.3389/fncom.2013.00127

Noble, K., Glowinski, D., Murphy, H., Jola, C., McAleer, P., Darshane, N., et al.(2014). Event segmentation and biological motion perception in watchingdance. Art Percept. 2, 59–74. doi: 10.1163/22134913-00002011

Pollick, F., Noble, K., Darshane, N., Murphy, H., Glowinski, D., McAleer, P.,et al. (2012). Using a novel motion index to study the neural basis of eventsegmentation. i-Perception 3, 225–225.

Sargent, J. Q., Zacks, J. M., Hambrick, D. Z., Zacks, R. T., Head, D., Kurby, C. A.,et al. (2013). Event segmentation ability uniquely predicts memory across thelifespan. Cognition 129, 241–255. doi: 10.1016/j.cognition.2013.07.002








Schachner, A., and Carey, S. (2013). Reasoning about ‘irrational’ actions: whenintentional movements cannot be explained, the movements themselves areseen as the goal. Cognition 129, 309–327. doi: 10.1016/j.cognition.2013.07.006

Schütz-Bosbach, S., and Prinz, W. (2007). Perceptual resonance: action-induced modulation of perception. Trends Cognit. Sci. 11, 349–355. doi:10.1016/j.tics.2007.06.005

Sevdalis, V., and Keller, P. E. (2011). Captured by motion: Dance, actionunderstanding, and social cognition. Brain Cogn. 77, 231–236. doi:10.1016/j.bandc.2011.08.005

Sridharan, D., Levitin, D. J., Chafe, C. H., Berger, J., and Menon, V. (2007).Neural dynamics of event segmentation in music: converging evidencefor dissociable ventral and dorsal networks. Neuron 55, 521–532. doi:10.1016/j.neuron.2007.07.003

Steggemann, Y., Engbert, K., and Weigelt, M. (2011). Selective effects of motorexpertise in mental body rotation tasks: comparing object-based and per-spective transformations. Brain Cogn. 76, 97–105. doi: 10.1016/j.bandc.2011.02.013

Stevens, C., Winskel, H., Howell, C., Vidal, L. M., Latimer, C., and Milne-Home,J. (2010). Perceiving dance: schematic expectations guide experts’ scanning of acontemporary dance film. J. Dance Med. Sci. 14, 19–25.

Swallow, K. M., Zacks, J. M., and Abrams, R. A. (2009). Event boundaries in per-ception affect memory encoding and updating. J. Exp. Psychol. Gen. 138, 236.doi: 10.1037/a0015631

WMA. (2008). “World Medical Association Declaration of Helsinki: ethical prin-ciples for medical research involving human subjects,” in 59th WMA GeneralAssembly (Seoul).

Waterhouse, E., Watts, R., and Bläsing, B. (2014). Doing Duo - a case study ofentrainment in William Forsythe’s choreography “Duo.” Front. Hum. Neurosci.8:812. doi: 10.3389/fnhum.2014.00812

Yarrow, K., Brown, P., and Krakauer, J. W. (2009). Inside the brain of an elite athlete:the neural processes that support high achievement in sports. Nat. Rev. Neurosci.10, 585–596. doi: 10.1038/nrn2672

Zacks, J. M., Kumar, S., Abrams, R. A., and Mehta, R. (2009). Using movementand intentions to understand human activity. Cognition 112, 201–216. doi:10.1016/j.cognition.2009.03.007

Zacks, J. M., Speer, N. K., Swallow, K. M., Braver, T. S., and Reynolds, J. R. (2007).Event perception: a mind-brain perspective. Psychol. Bull. 133, 273–293. doi:10.1037/0033-2909.133.2.273

Zacks, J. M., and Tversky, B. (2001). Event structure in perception and conception.Psychol. Bull. 127, 3–21. doi: 10.1037/0033-2909.127.1.3

Conflict of Interest Statement: The author declares that the research was con-ducted in the absence of any commercial or financial relationships that could beconstrued as a potential conflict of interest.

Received: 28 May 2014; accepted: 05 December 2014; published online: January2015.Citation: Bläsing BE (2015) Segmentation of dance movement: effects of expertise,visual familiarity, motor experience and music. Front. Psychol. 5:1500. doi: 10.3389/fpsyg.2014.01500This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2015 Bläsing. This is an open-access article distributed under the termsof the Creative Commons Attribution License (CC BY). The use, distribution or repro-duction in other forums is permitted, provided the original author(s) or licensor arecredited and that the original publication in this journal is cited, in accordance withaccepted academic practice. No use, distribution or reproduction is permitted whichdoes not comply with these terms.

www.frontiersin.org January 2015 | Volume 5 | Article 1500 |

07

74












Expertise affects representation structure and categoricalactivation of grasp postures in climbingBettina E. Bläsing1,2,3†, Iris Güldenpenning1,2*†, Dirk Koester1,2 and Thomas Schack1,2,3

1 Neurocognition and Action–Biomechanics Research Group, Faculty of Psychology and Sport Science, Bielefeld University, Bielefeld, Germany2 Center of Excellence-Cognitive Interaction Technology, Bielefeld University, Bielefeld, Germany3 Research Institute for Cognition and Robotics (CoR-Lab), Bielefeld University, Bielefeld, Germany

Edited by:


Reviewed by:

Markus Janczyk, University ofWürzburg, GermanyOliver Herbort,Julius-Maximilians-UniversitätWürzburg, Germany

*Correspondence:

Iris Güldenpenning, Neurocognitionand Action Research Group,Bielefeld University, PO Box 100131, 33501 Bielefeld, Germanye-mail: [email protected]†First authorship.

In indoor rock climbing, the perception of object properties and the adequate executionof grasping actions highly determine climbers’ performance. In two consecutiveexperiments, effects of climbing expertise on the cognitive activation of grasping actionsfollowing the presentation of climbing holds was investigated. Experiment 1 evaluated therepresentation of climbing holds in the long-term memory of climbers and non-climberswith the help of a psychometric measurement method. Within a hierarchical splittingprocedure subjects had to decide about the similarity of required grasping postures.For the group of climbers, representation structures corresponded clearly to four griptypes. In the group of non-climbers, representation structures differed more stronglythan in climbers and did not clearly refer to grip types. To learn about categoricalknowledge activation in Experiment 2, a priming paradigm was applied. Images of handsin grasping postures were presented as targets and images of congruent, neutral, orincongruent climbing holds were used as primes. Only in climbers, reaction times wereshorter and error rates were smaller for the congruent condition than for the incongruentcondition. The neutral condition resulted in intermediate performance. The findingssuggest that perception of climbing holds activates the commonly associated graspingpostures in climbers but not in non-climbers. The findings of this study give evidencethat the categorization of visually perceived objects is fundamentally influenced by thecognitive-motor potential for interaction, which depends on the observer’s experience andexpertise. Thus, motor expertise not only facilitates precise action perception, but alsobenefits the perception of action-relevant objects.

Keywords: expertise, mental representation, action activation, categorical knowledge, grasping

INTRODUCTIONRock climbing requires a multitude of physical and cognitive abil-ities, as well as their well concerted interaction. One of them isthe ability to perceive properties of climbing holds and to executeadequate grasping actions. In indoor climbing, the athlete’s goalis to reach the top of a climbing wall by using specific climbingholds that are arranged as routes of different skill requirements.The shape, orientation and relative position as well as the combi-nation of holds thereby determines the adequate grasp and steptechniques. Apprehending climbing holds correctly is crucial forplanning corresponding actions, and thereby for optimizing theclimber’s performance (Boschker et al., 2002). Thus, the ability toassign optimal grasps to the reachable holds is a relevant part of aclimber’s specific cognitive expertise (Pezzulo et al., 2010).

The cognitive aspects of expertise in the domain of rock orindoor climbing have hardly been investigated. One of the rareexperimental studies on cognitive issues of climbing expertise hasbeen conducted by Pezzulo et al. (2010), who investigated theability of novice and expert climbers to memorize climbing routesfrom presented photographs. Cognitive performance was mea-sured as number of recalled grips in the correct sequence of a

route. The main finding was that expert climbers outperformednovices in recalling the sequence of climbing holds of a difficultclimbing route that could only be climbed by experts, but not inrecalling an easy route that could be climbed both by experts andnovices. The authors argued that having the motor competenceto climb a visually perceived route enables climbers to mentallysimulate mastering it, and that this motor simulation improvesthe climbers’ recall of the grip sequence. Moreover, as partic-ipants were not explicitly instructed to mentally simulate therequired climbing actions, it was suggested that visually perceiv-ing a climbing route automatically activates the correspondingaction sequence in skilled climbers.

Automatic activation of motor components of graspingactions has also been found in other contexts, for example, whenparticipants had to classify kitchen objects or tools (Labeye et al.,2008), or manufactured and natural objects (Tucker and Ellis,2001, 2004; Grèzes et al., 2003). Labeye et al. (2008) investigatedthe processing of object features (kitchen tools vs. do-it-yourself(DIY) tools) and action features (implied actions performed witheach tool). Participants had to categorize target pictures as kitchentools or DIY tools. The targets were preceded by a prime picture
















Bläsing et al. Cognitive expertise in climbing

either depicting a kitchen tool or a DIY tool. Both the object(same vs. different category) and the action features (similarvs. dissimilar implied action) were independently manipulatedbetween the prime and the target pictures. Object and action fea-tures independently led to faster processing if they were from thesame category or implied similar actions, respectively, comparedto different categories or dissimilar actions. Labeye et al. (2008)argued that perceiving the prime picture automatically activatesthe motor components of the action that may potentially be per-formed with regard to the object (see also Ellis and Tucker, 2000;Tucker and Ellis, 2001; Bub and Masson, 2010; Masson et al.,2011).

Studies investigating well-known everyday objects (e.g.,kitchen tools) suggest that such object representations are asso-ciated with certain grasping actions (e.g., Tucker and Ellis, 2004;Labeye et al., 2008). These associations may emerge as a con-sequence of associated action experience. To further investigatethe role of motor experience for object-based action activations,laboratory training studies have been conducted (Creem-Regehret al., 2007; Kiefer et al., 2007; Weisberg et al., 2007; Cross et al.,2012; Bellebaum et al., 2013). After a training period in whichparticipants explicitly learnt to use objects in a tool-like man-ner, the manipulation experience became a part of participants’object representations and were automatically activated when theobjects were perceived (e.g., Weisberg et al., 2007).

A sports context provides a suitable scenario for investigat-ing expertise effects in object-related action knowledge. Climbingholds have been artificially designed to be used in indoor climb-ing and are encountered exclusively in this context. Accordingly,people who do not practice indoor climbing have no manipula-tion experience with climbing holds, whereas sport climbers whofrequently train on indoor climbing walls have a large amount ofspecific manipulation experience. Yet, climbing holds are objectsfrom which particular manipulation potential might be inferredeven by non-climbers based on the perceivable shape-propertiesof the object. Comparing non-climbers’ and climbers’ processingof climbing holds therefore provides a suitable scenario to dissoci-ate the role of grasping experience and physical object propertiesin the representation and processing of grasping actions.

Climbing is not a sport performed under time pressure (exceptspeed climbing; Florine and Wright, 2004). However, automaticactivations of single grasping actions are an important aspect ofa climber’s performance. The immediate activation of a grasp-ing action to a perceived climbing hold might decrease cognitiveeffort in short-term memory and thus save cognitive resourcesnecessary for further action planning (Spiegel et al., 2013).Besides, the direct activation of a grasping action also allowsa quick action execution and may thus prevent high energycosts arising when a climber has to remain in a static positionevaluating the next action possibility. Based on these considera-tions, the present study examines object-related action knowledge(Experiment 1) and automatic action activation (Experiment 2)based on perceived climbing holds.

To investigate knowledge representations of climbing-specificgrasping actions, Experiment 1 evaluated the relevant cogni-tive structures of grasping actions related to typical climbingholds in the long-term memory of climbers and non-climbers

via Structure Dimensional Analysis (Schack, 2004; SDA; Schack,2012). It was expected that climbers, but not non-climbers,would categorize climbing holds according to appropriate griptypes (functional features) used in indoor climbing rather thanaccording to other object properties.

Experiment 2 was conducted to clarify whether or not the rep-resentational clusters are actually associated with motor compo-nents that fit the holds in climbers. A priming paradigm was usedwith pictures of climbing holds as primes and grasping posturesas targets (e.g., Güldenpenning et al., 2011, 2013). Climbers wereexpected to show different effects than non-climbers. Specifically,climbers but not non-climbers should show a congruency effect,i.e., facilitation by congruent primes and inhibition by incongru-ent primes (Dehaene et al., 1998).

EXPERIMENT 1: CATEGORIZATION OF CLIMBING HOLDSIn indoor climbing, climbers have to manage routes of differentdifficulty level, which requires a multitude of physical and cog-nitive skills. Applying adequate grasp techniques to the availableclimbing holds is one of the crucial tasks in this challenge, as itenables the climber to master the route in a safe and efficientway. Experiment 1 investigated if climbing holds are categoricallyorganized depending on their associated manual actions (i.e.,adequate grip types) in the long-term memory of experiencedindoor climbers. Structure dimensional analysis (SDA; Schack,2004, 2012) was applied to reveal the relevant representationalstructures related in the long-term memory of climbers and non-climbers. It was expected that climbers, but not non-climbers,categorized climbing holds according to the appropriate griptypes. Previous studies using SDA showed that experts’ clustersolutions referred to functionally structured mental representa-tions of complex movements (Bläsing et al., 2009; Land et al.,2013) or to objects affording similar actions (e.g., Stöckel et al.,2012) representing functional action-based categories of partialactions or objects.

METHODSParticipantsTwenty-one participants voluntarily took part in Experiment 1without any exchange or in exchange for course credit. Tenstudents of sport science without any experience in indoor oroutdoor rock climbing, all from Bielefeld University, Germany,were assigned to the non-climbers’ group (two females, all right-handed, mean age 24.0 years, range 23–25). All non-climberswere physically active (eight out of ten participants performedindividual sports, four performed team sports). The sports mostregularly performed by the participants of the non-athlete groupincluded soccer, basketball, fitness training, and running.

Eleven climbers (two females, all right-handed, mean age 27;range 22–34) were recruited from an indoor climbing area, dueto their experiences in climbing (mean climbing experience: 5.3years of training, 3.4 training sessions per week). Referring tothe Union Internationale des Associationes d’Alpinisme’s (UIAA)grading system describing the difficulty of the climbing route,participants’ indoor climbing skills ranged from 6 to 8 (two par-ticipants climbed routes graded up to 8, five participants climbedroutes graded 7). Five of the climbers also regularly climbed






outdoors, with skills corresponding to routes ranging from 5 to 7.Additionally, five out of the ten climbers regularly performedother sports (mostly mountain sports or running).

All participants reported having normal or corrected-to-normal vision, and were naive with regard to the purpose ofthe experiment. All participants provided written informed con-sent before testing started. The single experimental session lastedabout 30 min. The experiment was performed in accordance withthe ethical standards of the sixth revision of the 1964 Declarationof Helsinki (World Medical Assocition, 2008).

Apparatus and stimuliStimuli consisted of 16 colored photographs of climbing holds ofdifferent shapes and sizes, as commonly used in indoor climbing(see Figure 1). The holds were presented to match the climber’sperspective, in adequate size relative to each other. All holds werechosen to elicit specific grip types rather than being ambiguous ornon-specific. Six out of sixteen holds typically required a crimpgrip, four a sideways pull toward the body, four a pocket grip, andtwo an open grip. This a priori attribution of climbing holds togrip types was informed by climbing experts who did not partic-ipate in the study and was used as reference for the results of theexperiment.

Design and procedureThe participants were tested individually while sitting in front ofa computer screen. An experimental paradigm named StructureDimensional Analysis (SDA; Schack, 2004, 2012) was applied toinvestigate the categorization of climbing holds on the basis of

mental representations of specialist grip types in the long-termmemory of climbers and non-climbers. SDA was applied viacustom-made software (NetSplit). The stimuli were presented insuch a way that in each trial, the reference stimulus (or anchor)occurred in the top position marked by a green frame, and thestimulus directly below the reference, marked by a blue frame, hadto be assigned by key press to a positive (left) or negative (right)list relative to the reference (see Figure 2). The anchor and theactive stimulus picture were presented on the screen with a size ofapproximately 6 × 6 cm. The participants were instructed to indi-cate by pressing one of two marked keys if the adequate graspingaction directed toward the currently active hold (marked blue)would be of the same type as the one directed toward the climbinghold in the reference position (marked green). Once the responsewas given, the next trial began, in which the same anchor was pre-sented with another of the remaining items, until all 15 items hadbeen judged as affording a similar or dissimilar grip compared tothe anchor; this procedure composed one block. In the next block,a different hold was presented as anchor, in combination with allremaining 15 holds. The whole experiment comprised 16 blocksapplied in randomized order, each block with a different item asanchor, resulting in a total of 240 trials.

Data analysesA hierarchical cluster analysis according to SDA (Lander andLange, 1996; Schack, 2004, 2012) was carried out on the data col-lected via the previously described splitting procedure in order toobtain mean cluster solutions for the two experimental groups. Toachieve this, the sorting task described as part of the experimental

FIGURE 1 | Stimulus pictures showing climbing holds commonly used in indoor climbing; this grouping of climbing holds according to grip types

was used as reference for the participants’ results.





FIGURE 2 | Presentation of stimuli during the experiment. Thestimulus item in the top position, marked by a green frame, is thecurrent anchor (reference) relative to which the active stimulusbelow, marked by a blue frame, has to be categorized as affording

a similar grip or affording a dissimilar grip. Items appearing insmaller size left and right of the active item have already beenassigned to the positive list (similar grips) or to the negative list(dissimilar grips), respectively.

procedure was applied to deliver a distance scaling between theitems (climbing holds). By this procedure, 16 decision trees wereestablished, as each item occupied once the reference position,resulting in a 16 × 15 matrix of partial quantities in which val-ues took either a negative or positive sign depending on whetherthe item was judged as belonging to the positive or the neg-ative list relative to the anchor (e.g., if 4 out of the 15 itemswere assigned to the positive list, these items were each given thevalue +4, whereas the remaining 11 features assigned to the nega-tive list were each given the value −11). The resulting matrix wasthen z-transformed for standardization and subsequently trans-formed into a Euclidian distance matrix as basis for a hierarchicalcluster analysis (in accordance with the average-linkage-method).Cluster solutions were determined using a critical Euclidian dis-tance (dcrit), with all junctures lying below this value forming theapical pole of an underlying concept cluster (for more details onthe method, see Schack, 2012). As reference structure, an a pri-ori classification of stimulus climbing holds according to specificgrip types had been achieved based on interviews with climb-ing experts who did not participate in the study (see Figure 1).To calculate the similarity between mean group results with thereference structure and to compare each individual participant’scluster solution with the averaged cluster solution of the groupand the reference structure (holds 1–6: crimp grip; holds 7–10:sideways pull; holds 11–14: pocket grip; holds 15 and 16: opengrip; see also Figure 1), we used the adjusted rand index (ARI;Rand, 1971; Santos and Embrechts, 2009). The ARI provides ameasure of similarity on a range between 0 and 1; a score of

0 indicates that two cluster solutions are independent, whereasa score of 1 indicates that two cluster solutions are identical.Scores between these two values indicate the degree of similar-ity between cluster solutions; the higher the ARI score, the greateris the similarity between the variables.

ResultsThe hierarchical cluster analysis revealed four clusters corre-sponding to four grip types for the group of climbers, and threesmaller clusters for the non-climbers. The four clusters of theclimber group included all 16 holds into clusters that reflectedthe correct assignment of holds to grip types (cluster 1: items 1–6,cluster 2: 7–10, cluster 3: 11–14, cluster 4: 15 and 16). Euclideandistances between the items of all clusters were all below 1.5 (criti-cal value: 3.4, alpha value: 5%), which reflected a high consistencyof the climbers’ decisions. The non-climbers’ cluster solution con-tained three clusters, consisting of items 2 to 6, 7 and 8, and 13 and14. Euclidean distances were all larger than 1.5, and the remainingseven items were singled out (i.e., were not significantly assignedto any cluster). The mean group dendrograms for climbers andnon-climbers are presented in Figure 3.

To compare individual participants’ cluster solutions to thegroups’ mean cluster solutions, adjusted rand index (ARI, Santosand Embrechts, 2009) was calculated, which expresses the extentto which the individual cluster solutions differ from the respec-tive averaged group dendrogram. Comparison of the mean groupcluster solutions with the reference structure resulted in an ARIscore of 1.0 for the climbers (i.e., both cluster solutions were






FIGURE 3 | Results of Experiment 1: mean cluster solutions of groups;

top: climbers (N = 11); bottom: non-climbers (N = 10). Left: dendrograms;alpha: 5%, dcrit (marked by a horizontal line): 3.4. Numbers on the x-axis refer to

the item numbers; numbers on the y-axis indicate Euclidean distance betweenitems according to the cluster analysis via SDA. Right: Stimulus picturesrepresenting cluster solutions.

identical) and of 0.535 for the non-climbers. Individual ARIscores of the climbers ranged from 0.685 to 1.0, with a meanof 0.952 ± 0.195 (SD). For the non-climbers, ARI scores weresmaller than the climbers’ (Mann-Whitney U-test; Z = −3.887,p < 0.001), they ranged from 0.0 to 0.643, with a mean of0.311 ± 0.206. When non-climbers’ ARI scores were calculatedwith reference to the reference structure, they were also smallerthan the climbers’ (Z = −4.033, p < 0.001), ranging from 0.0to 0.638, with a mean of 0.268 ± 0.194. The non-climbers’ ARIscores calculated relative to the group average and the referencestructure did not differ (Wilcoxon signed-rank test; Z = −1.836,p = 0.066).

Discussion experiment 1The hierarchical cluster analysis via SDA revealed four clusters forthe group of climbers, and three clusters for the group of non-climbers. For the group of climbers, the mean cluster solutionwas identical with the functional assignment of grasping holdsto grip types (see Figure 1), and individual participants’ clustersolutions differed only little from each other; nine out of elevenparticipants produced a result that was identical with the meancluster solution. Euclidean distances between items within eachcluster were all below 1.5, and thereby small compared to the

critical value (dcrit = 3.4). These results point toward a high con-sistency of decisions made by participants during the experiment,within as well as between participants. These findings suggeststhat climbers, on the basis of their experience in indoor climb-ing, could easily associate the presented climbing holds to thecorresponding grip types, thereby producing consistent clustersrepresenting functional task-related categories.

In contrast, the non-climbers’ mean cluster solution includedonly nine out of 16 items into clusters, whereas the remainingseven items remained as singletons (that is, these seven itemswere not categorized by the non-climbers, reflecting a partly non-categorical representation). The three clusters each containeditems that belonged to the same grip category. This finding couldbe explained by two mechanisms: despite their lack of climbingexperience, non-climbers might have been able to assign certainclimbing holds to appropriate grip types, potentially based ontheir experience with manipulable objects from a non-climbingcontext. Non-climbers thereby might have applied the same (orsimilar) criteria as climbers, but succeeded in doing so only fora subset of the presented items. In this case, the results reflectthat attribution to a specific grip type was more difficult for cer-tain climbing holds than for others (e.g., the appropriate grip thatrequired inserting one or more fingers into openings in the hold





was apparently easier to recognize for items 13 and 14 than foritems 11 and 12).

Alternatively, non-climbers might have grouped items on thebasis of other feature-based object similarities related to shape oreven color. The latter explanation is supported by the observa-tion that items grouped into the same cluster by the non-climberslooked similar (this is particularly obvious for items 7 and 8 andfor items 13 and 14, respectively). Previous studies have shownthat novice participants often tend to combine items that showsimilarity regarding superficial characteristics, rather than task-related functional dependence, into the same cluster (see Schack,2004). In studies that use partial movements (or basic action con-cepts, see e.g., Schack and Mechsner, 2006; Schack and Ritter,2009) in order to investigate mental representations of complexmovements, such characteristics often regard the use of bodyparts (e.g., movement concepts referring to the arms might becombined into one separate cluster, even though this might haveno relevance for the functional structure of the movement, asarm movements might have different and even opposed functionsduring different movement phases). The results suggest that thenon-climbers, due to their lack of task-related experience, wereless able than the climbers to decide consistently which specificgrips were required for the presented climbing holds.

These findings corroborate the notion that the categorizationof visually perceived objects is fundamentally influenced by theirpotential for interaction (i.e., their affordances; Gibson, 1977),the evaluation of which strongly depends on the observer’s task-specific experience and expertise. For task-specific objects suchas climbing holds, certain features determine their potential useand are, therefore, relevant for functional categorization, whereasother features play a minor role. Climbers, compared to non-climbers, choose more purposefully which of the characteristicsof a climbing hold are relevant for the task in question. (In thecurrent study, shape and orientation were task-relevant features,whereas in a different context, the color of climbing holds couldbe a task-relevant object feature, as it would allow the climber toview the hold as part of a color-coded route).

Taken together, the results of Experiment 1 show that thecognitive representation of objects (i.e., climbing holds) stronglydepends on their functional relation to action-based experi-ences. Based on this finding, we investigated in the subsequentexperiment with similar stimuli and two similar groups of partic-ipants if and to what extent the perception of task-related objectsinfluences (short-term) processing and the (pre-)activation ofobject-related actions.

EXPERIMENT 2: COGNITIVE ACTIVATION OF GRASPINGPOSTURESExperiment 2 investigated whether visually perceiving a climb-ing hold activates the grasping posture commonly associated withthis climbing hold. Moreover, it was asked whether the activationof the grasping postures depends on the manipulation experi-ences with the climbing holds. In a priming experiment, climbersand non-climbers were asked to decide whether a presented targetpicture reflected a crimp grip or a sideways pull. The preced-ing prime picture either depicted a climbing hold requiring thegrip shown in the target picture (congruent condition; e.g., a

hold requiring a sideways pull followed by a sideways pull) or thealternative grip (incongruent condition; e.g., a hold requiring asideways pull followed by a crimp grip). Moreover, two unspe-cific conditions were applied. In the positive unspecific condition,the target picture was preceded by a climbing hold which couldboth be grasped with a crimp grip as well as a sideways pull. Inthe negative unspecific condition, the preceding climbing holdcould neither be grasped with a crimp grip nor with a sidewayspull.

It was expected that in experienced climbers a climbing holdwould activate the grasping posture commonly associated withthe climbing hold, and thus influence responses to the tar-get picture depicting a particular grasping posture (i.e., crimpgrip vs. sideways pull). No such activation was expected in par-ticipants without manipulation experience with the climbingholds. Moreover, for both groups it was expected that unspecificclimbing holds would not activate any grasping posture.

The following three predictions were made: first, a congruencyeffect was expected for participants with climbing experience; thatis, faster response times under conditions in which a climbinghold shown in the prime picture would require the same grasp-ing posture as depicted in the following target picture, and slowerresponse times under conditions in which a climbing hold wouldrequire the alternative grasping posture as shown in the targetpicture. Second, no congruency effect was expected for partici-pants without specific climbing experience. Third, for climbers,response latencies in the unspecific condition were predicted to bein between the congruent and the incongruent condition, whereasfor non-climbers, response times should not vary between con-ditions. The described differences between groups related to thefactor congruency are expected to be statistically indicated by aninteraction between group and congruency.

METHODSParticipantsThirty two participants voluntarily took part without anyexchange or in exchange for course credit. Eighteen students oremployees from Bielefeld University, Germany, were assigned tothe non-climber group (three female, one left-handed, mean age29.6; range 23–43). Participants of the non-climber group had noexperience in climbing. All non-climbers were physically active,performing at least one type of sport (11 participants performedindividual sports, 12 performed team sports, 2 performed com-petitive sports, 2 performed racket sport). Participants of thenon-climber group played, for example, soccer, handball, bas-ketball, or regularly performed swimming, running, or fitnesstraining.

Fourteen climbers (one female, two left-handed, mean age30.0; range 16–43) were recruited from an indoor climbingarea for their experiences in climbing (mean training expe-rience: 8.7 years; mean training frequency per week: 4.0 ses-sions). All recruited climbers were members of the DeutscherAlpenverein (German Alpine Association). Referring to the UIAAgrading system describing the difficulty of climbing routes, par-ticipants’ climbing skills ranged between 6 and 10 (two par-ticipants were able to climb routes graded 6, two participantsclimbed routes graded 7, six participants climbed routes graded






8, three participants climbed routes graded 9, and one participantclimbed routes graded 10).

All participants reported to have normal or corrected-to-normal vision and were naive with regard to the purpose ofthe experiment. All participants provided written informed con-sent before testing started. The single experimental session lastedabout 20 min. The experiment was performed in accordance withthe ethical standards of the sixth revision of the 1964 Declarationof Helsinki (World Medical Assocition, 2008).

Apparatus and stimuliFor stimulus presentation and data collection, a ToshibaNotebook with a 17 inch VGA-Display (vertical retraces 60 Hz)and the software Presentation® (version 14.8) was used. The soft-ware controlled the presentation of the stimuli and measuredreaction times. Responses had to be given by pressing one of twoexternal response buttons connected via a parallel port with thenotebook.

The target pictures were 16 photographs of hands in graspingpostures. Eight pictures reflected a crimp grip and eight picturesreflected a sideways pull. Half of the grasping postures reflected aright hand and half of the postures reflected a left hand (imageswere mirrored) which were used equally often.

As prime pictures, 32 photographs of climbing holds werepresented. All climbing holds were red. Four of the climbingholds would require a crimp grip, and four climbing holds wouldrequire a sideways pull. Moreover, four climbing holds could begrasped with a crimp grip or a sideways pull (positive unspecificclimbing holds). Last, four climbing holds would require a gripthat was different from a crimp grip or a sideways pull (negativeunspecific climbing hold). The 16 pictures of the climbing holdswere mirrored vertically to extend the spectrum of the stimulusmaterial to 32 prime pictures in total. An exemplary illustrationof the stimulus material is given in Figure 4.

The background of the climbing holds and of the graspingpostures was an indoor climbing wall (see Figure 4). The stim-uli had a size of 9.2 × 9.2 cm (346 × 346 pixels). All stimuli werepresented centrally on a black background and subtended a visualangle of 8.9◦ horizontally and vertically from the viewing distanceof 60 cm.

Design and procedureThe present study used a 4 × 2 mixed factorial design with thewithin-subject factor congruency (congruent condition, incon-gruent condition, positive and negative unspecific condition)and participants’ expertise as between-subject factor (climbers vs.non-climbers). The impact of these factors was analyzed withreaction time (RT) and error rate (ER) measures as the dependentvariables.

Participants sat in front of a computer screen (60 cm) andwere instructed in written form to classify the presented tar-get picture as a crimp grip or as a sideways pull as quickly aspossible by pressing one of the two response buttons with theindex finger. Moreover, participants were instructed to respondas accurately as possible. The response button assignment wascounterbalanced across participants within each group. Beforestarting the experimental session, each participant performed tenrandomized practice trials. Data from this practice block werenot analyzed. The following test block consisted of 128 pseudo-randomized prime-target pairs. Each prime picture appeared fourtimes and was either combined with a left hand or right handcrimp grip or with a left hand or right hand sideways pull. Thepresentation of the order of the prime-target pairs was completelyrandomized.

Each trial started with the presentation of a central fixa-tion cross (400 ms), followed by a blank screen (100 ms), theprime (100 ms), a second blank screen (100 ms), and the tar-get (which remained visible on the screen until a response was

FIGURE 4 | Examples of the stimulus material used in Experiment 2. On the left side examples for each type of climbing hold are given. On the right sideexamples for each type of grasping posture are presented.





given). Incorrect responses elicited the word “Fehler” (Germanfor “error”). An inter-trial interval of 1500 ms elapsed beforethe next trial started. The within-trial procedure is illustrated inFigure 5.

RESULTSData analysesReaction times (RTs) were screened for outliers using a total cutoff. RTs below 200 ms and above 1000 ms were excluded (2.0%).Trials with wrong answers (3.4%) were not used in the analy-sis of the RTs. The mean RTs from the factorial combinations ofthe within-subjects factor congruency and the between-subjectsfactor expertise were computed for further analysis. A prelimi-nary comparison of the RTs for positive unspecific primes andnegative unspecific primes was performed. Separate paired t-tests revealed neither a significant difference between positive(547 ms, s.e.m. = 21 ms) and negative unspecific primes (552 ms,s.e.m. = 22 ms) for climbers, t(13) = 0.56, p = 0.59, nor for pos-itive (529 ms, s.e.m. = 14 ms) and negative unspecific primes(535 ms, s.e.m. = 14 ms) for non-climbers, t(17) = 0.98, p =0.34. This result indicates that positive unspecific primes and neg-ative unspecific primes did not evoke differential priming effects.Thus, further analyses were computed with the mean value ofpositive unspecific and negative unspecific primes. This param-eter value of the factor congruency was simply termed neutralcondition.

Mixed ANOVAs with the within-subjects factor congruency(congruent, incongruent, neutral) and the between-subjects fac-tor expertise (climbers vs. non-climbers) were performed withRT and ER as dependent variables. A violation of the sphericity-assumption resulted in a correction of the p-values according toGreenhouse-Geisser1.

Reaction timesThe within subjects factor congruency reached significance,F(1, 60) = 4.86, p = 0.01, ε = 0.84, η2

p = 0.14, as well as the

1A violation of the sphericity-assumption requires modifications of thedegrees of freedom (df s) to reduce the Type I error. The estimates for thismodification is denoted by epsilon (ε) providing a measure of deviation fromsphericity. An epsilon of 1 indicates that sphericity is exactly met, smallerepsilon values indicate increasing deviation. Due to better reading, the uncor-rected df s are given for the F-value, but the corrected df s can be easilycalculated by multiplying the original df s with ε.

interaction between congruency and expertise, F(2, 60) = 3.88, p =0.03, ε = 0.84, η2

p = 0.12. The between subjects factor expertisedid not reach significance (p = 0.75).

To illuminate the source of the interaction, paired t-tests wereperformed separately for climbers (one-tailed) and non-climbers(two-tailed). Climbers responded significantly faster to congru-ent compared to incongruent prime-target pairs, t(13) = 2.52,p = 0.01. Responses to grasping postures preceded by a neu-tral prime were significantly slower than responses to congruentprime-target pairs, t(13) = 2.02, p = 0.03, and significantly fasterthan responses to incongruent prime-target pairs, t(13) = 1.89,p = 0.04.

Non-climbers in contrast responded not differently fast (p =0.72) to grasping postures preceded by a congruent climbing holdand by an incongruent climbing hold. Interestingly, responsesto grasping postures preceded by a neutral climbing holdwere significantly faster than congruent, t(17) = 2.18, p = 0.04,and also faster than incongruent climbing holds, t(17) = 2.78,p = 0.01.

Mean values of the RTs and ERs and corresponding confidenceintervals are illustrated in Figure 6 and additionally displayed inData Sheet 1 in the Supplementary Material.

Response errorsA mixed ANOVA on the mean ERs neither revealed a significanteffect for the within subjects factor congruency (p = 0.23) nor forthe between subjects factor expertise (p = 0.60). The interactionbetween congruency and expertise reached statistical significance,F(2, 60) = 4.30, p = 0.01, ε = 0.82, η2

p = 0.17. To compare theresults of the analysis of the RTs with the ERs, paired t-tests werecomputed separately for climbers (one-tailed) and non-climbers(two-tailed).

For climbers, responding was less error-prone with congruentcompared to incongruent trials, t(13) = 2.80, p = 0.01. The com-parison between the incongruent and the neutral condition alsoreached statistical significance, t(13) = 2.8, p = 0.01, indicatinga higher ER for incongruent compared to neutral prime-targetpairs. The comparison between the congruent and the neutralcondition did not reach significance (p = 0.27).

For non-climbers, none of the comparisons revealed a signif-icant effect (all ps > 0.30), but the error rate was slightly smallerfor incongruent primes compared to congruent and neutralprimes.

FIGURE 5 | Stimulus presentation in Experiment 2. This figure reflects the within-trial procedure for an example of the incongruent condition: a climbinghold requiring a crimp grip is followed by a target picture depicting a sideways pull.






FIGURE 6 | Results of Experiment 2: Reaction times (RTs) in milliseconds

with corresponding 95% confidence intervals (CIs) and error rates (ERs)

in percent with corresponding 95% CIs for climbers (left panel) and

non-climbers (right panel). The dashed lines display RTs scaled on the leftvertical axis. The bars illustrate ERs scaled on the right vertical axis. Theprime-target conditions are labeled on the x-axis.

Discussion experiment 2Experiment 2 aimed to investigate the activation of graspingpostures by visually presenting climbing holds and how such acti-vation is influenced by skill level. In accordance with the hypothe-ses, a congruency effect was found for climbers, that is, fasterresponses were found for congruent trials compared to incongru-ent trials. The inclusion of an unspecific condition revealed thatthe found congruency effect is based both on speeded responsesin the congruent condition and on slowed responses in the incon-gruent condition compared to the neutral baseline. It would beinteresting to determine whether this priming effect arises atthe perceptual level (perceptual priming, e.g., Biederman andCooper, 1992), the cognitive level (e.g., Labeye et al., 2008), orat a motor stage of processing (response priming; e.g., Kundeet al., 2003). Regarding a perceptual locus of the priming effects,there is no visual relation or similarity between the climbingholds and the grasping postures that is larger for the congruentprime target pairs (hold-posture) compared to the incongruentpairs. Hence, we consider it implausible that our priming effectsare due to perceptual (dis)similarity of the stimuli. Regarding apotential motor locus, the priming effect is unlikely to reflect aresponse activation or a response competition effect because thetask instruction to classify a grasping posture as a crimp grip oras a sideways pull cannot directly be applied to the climbing holdpictures (primes). That is, the holds themselves should not acti-vate or elicit a response per se (i.e., pressing a left or right responsebutton). Therefore, we would argue that the priming effects donot arise during perceptual or motor stages but during cognitiveprocessing stages.

More precisely, the result pattern of speeded responses in con-gruent trials and slowed responses in incongruent trials, bothrelative to a neutral baseline, points toward two cognitive pro-cessing mechanisms. First, the perception of a given hold leadsto an activation of the representation of that hold including thecorresponding grasping action. Second, the perception of a givenhold seems to lead to a reduced availability of non-correspondinggrasping actions. These mechanisms of action activation and

action inhibition might help to explain the efficient selection ofgrasping actions in climbing.

The results found in the non-climbers are also in accordancewith the interpretation of a cognitive locus of the priming effects.Participants with no climbing experience are expected to pos-sess no representations of climbing-specific grasping actions.Accordingly, non-climbers did not show any congruency effect.Unexpectedly, however, non-climbers had shorter reaction timesfor unspecific prime-target pairs compared to congruent undincongruent ones. Besides the possibility that this finding mightbe a random finding in principle, a possible post-hoc explana-tion for this result could be the following. Round objects similarto the unspecific climbing holds are also common in daily life,for example, as round rotatable button on a washing machineor as knob-like hold of a drawer. It is thus possible that non-climbers applied their general knowledge, that round objectscan be grasped either with a crimp grip or with a sidewayspull (without knowing these climbing specific concepts), to thepredominantly round shapes of climbing holds in the neutralcondition. In contrast, non-climbers could not infer any associ-ated grasping action from the unfamiliar climbing holds requiringa crimp grip or a sideways pull. The faster responses to targetsfollowing a prime picture with a round hold thus may reflectunspecific activations of grasping representations. Further studiesare needed to confirm this suggestion.

GENERAL DISCUSSIONThis study explored the relationship between the action-basedcognitive representations of climbing holds and the object-basedactivation of the corresponding grasping actions. Experiment 1investigated the structure of skill representations. (Note that weuse the term climbing skill in this context specifically for theskilled manual use of climbing holds, i.e., the knowledge of cor-rect grip application for individual climbing holds.) The resultsof Experiment 1 suggest that climbers organize visually perceivedclimbing holds categorically according to functional features (i.e.,how to grasp in an indoor climbing context). Non-climbers, in





contrast, showed an organizational structure that was not cate-gorical in terms of skill-based (climbing) knowledge but ratherbased on unspecific world knowledge or superficial object fea-tures (e.g., form or color). Experiment 2 investigated the accessof grasping knowledge, in particular whether and how functionalfeatures (i.e., grasping postures) are activated by object features(i.e., shapes of climbing holds). Here, we found evidence foractivation of matching grasping postures and inhibition of non-matching grasping postures by the perception of grip-specificholds.

We argue that in climbers, but not in non-climbers, the cat-egorical memory structure reflects the functional features ofclimbing holds in the context of indoor climbing. This struc-ture appears to follow functional distinctions of the associatedactions and reflects climbers’ manifold experiences of task-relatedaction-effect coupling (Hoffmann, 2003). The qualitative changesin memory structures might also change perceptual informationprocessing (cf. Ericsson and Kintsch, 1995). Having commandover, for example, a grip category related to sideways pull wouldfacilitate the recognition and processing of potentially distinctholds regarding their applicability and functional relevance forthe adequate motor action. The present findings thus provideinsights as to how skilled climbers may achieve a better climbingperformance: automatic activation of adequate grasping actionsin response to the perception of a specific climbing hold can beregarded as crucial mechanism to reduce the cognitive demandsinvolved in decision making and the planning of selected motoractions. This mechanism thereby serves to reduce cognitive pro-cessing time, which, importantly in the climbing context, leads toreduced physical energy costs (i.e., muscle force needed to remainin a static posture while evaluating the next move). Climbingthereby represents a relevant example of how skill-based knowl-edge that can be accessed explicitly for cognitive control alsosupports evaluative, strategic action planning under resourceconstrained conditions.

Our results are in accordance with findings by Pezzulo et al.(2010) who reported better recall performance for difficult climb-ing routes in climbers compared with non-climbers. Pezzulo et al.(2010, p. 72) speculate that experts are better able to form motorchunks that are necessary for mastering perceived climbing routesby means of simulation and hence are better in memorizing theroutes, but these authors also consider alternative explanationssuch as better visual imagery in experts. Whereas Pezzulo et al.(2010) used an offline measure of memorization, Experiment2 in our study investigated online processing (i.e., short-termactivation) of skill representations which are suggested to becategorically structured according to our Experiment 1.

The assumption of categorical skill representations raises thequestion of how specific actions are selected. Generally, it is con-ceivable that some holds have multiple grasping possibilities andshould, thus, activate more than one category of climbing holds.Here, Experiment 2 yielded evidence for processes of activationand inhibition. The results of Experiment 2 are also in accordancewith Labeye et al. (2008) who also found that the perception of(manipulable) objects activates associated actions, even thoughwe used considerably longer stimulus onset asynchronies com-pared to Labeye et al. (2008). Moreover, Experiment 2 suggests

that such action feature activations depend on previous learningor experience with the object and not on the pure physical objectproperties as the non-climbers’ data pattern showed the fastestresponses in the unspecific conditions.

Our results corroborate findings from studies of complexmovement representations (Bläsing et al., 2009; Güldenpenninget al., 2011, 2013; Weigelt et al., 2011; Land et al., 2013) andskill acquisition (Frank et al., 2013). They emphasize the role ofcognitive representations and processes in action control and sup-port the view that skill representations are based on categoricalknowledge. In this regard, our study confirms the role of cognitiveprocesses in the control of complex human actions as proposedin frameworks such as the ideomotor approach (Koch et al., 2004;for a histoical overview of the ideo-motor principle, see Stock andStock, 2004), the theory of event coding (TEC; Hommel et al.,2001), or the cognitive action architecture approach (CAA-A;Schack and Ritter, 2009; Land et al., 2013).

Taken together, the present studies investigated the cognitiverepresentations of indoor climbing holds and the perceptual pro-cessing of such holds and associated grasping postures in climbersand non-climbers. Experienced climbers represent holds accord-ing to their functions (i.e., grip types) whereas non-climbers showless structure in their representations and organize these accord-ing to unspecific action knowledge or superficial features. It wasalso found that the perception of climbing holds activates thematching grasping posture and inhibits non-matching posturesin climbers but not in non-climbers. These findings suggest aprocessing advantage and mechanism of categorical action repre-sentations. Furthermore, the findings show that action experiencemodifies the relevant object representations by associating actionfeatures to the representations of corresponding objects.

ACKNOWLEDGMENTSThe authors thank Marco Schweizer and Chistiane Preuß forhelp with the production of stimulus pictures and acquisitionof data and Dr. Christoph Schütz for providing the NetSplitsoftware used in Experiment 1. This study was supported byGerman Research Foundation Grant DFG EXC 277 “CognitiveInteraction Technology” (CITEC) (www.cit-ec.de/). We acknowl-edge support for the Article Processing Charge by the DeutscheForschungsgemeinschaft (DFG) and the Open Access PublicationFunds of Bielefeld University Library.


REFERENCESBellebaum, C., Tettamanti, M., Marchetta, E., Della Rosa, P., Rizzo, G., Daum, I.,

et al. (2013). Neural representations of unfamiliar objects are modulated by sen-sorimotor experience. Cortex 49, 1110–1125. doi: 10.1016/j.cortex.2012.03.023

Biederman, I., and Cooper, E. E. (1992). Size invariance in visual object prim-ing. J. Exp. Psychol. Hum. Percept. Perform. 18, 585–593. doi: 10.1037/0096-1523.18.1.121

Bläsing, B., Tenenbaum, G., and Schack, T. (2009). The cognitive structureof movements in classical dance. Psychol. Sport Exerc. 10, 350–360. doi:10.1016/j.psychsport.2008.10.001


http://www.cit-ec.de/







Boschker, M. S., Bakker, F. C., and Michaels, C. F. (2002). Memory for the func-tional characteristics of climbing walls: perceiving affordances. J. Motor Behav.34, 25–36. doi: 10.1080/00222890209601928

Bub, D. N., and Masson, M. E. (2010). Grasping beer mugs: on the dynamics ofalignment effects induced by handled objects. J. Exp. Psychol. Hum. Percept.Perform. 36, 341–358. doi: 10.1037/a0017606

Creem-Regehr, S. H., Dilda, V., Vicchrilli, A. E., Federer, F., and Lee, J. N. (2007).The influence of complex action knowledge on representations of novel gras-pable objects: evidence from functional magnetic resonance imaging. J. Int.Neuropsychol. Soc. 13, 1009–1020. doi: 10.1017/S1355617707071093

Cross, E. S., Cohen, N. R., Hamilton, A. F., d,. C., Ramsey, R., Wolford, G., andGrafton, S. T. (2012). Physical experience leads to enhanced object perceptionin parietal cortex: insights from knot tying. Neuropsychologia 50, 3207–3217.doi: 10.1016/j.neuropsychologia.2012.09.028

Dehaene, S., Naccache, L., Le Clec’H, G., Koechlin, E., Mueller, M., Dehaene-Lambertz, G., et al. (1998). Imaging unconscious semantic priming. Nature 395,597–600. doi: 10.1038/26967

Ellis, R., and Tucker, M. (2000). Micro-affordance: the potentiation of compo-nents of action by seen objects. Br. J. Psychol. 91, 451–471. doi: 10.1348/000712600161934

Ericsson, K. A., and Kintsch, W. (1995). Long-term working memory. Psychol. Rev.102, 211–245. doi: 10.1037/0033-295X.102.2.211

Florine, H., and Wright, B. (2004). Speed Climbing!: How to Climb Faster and Better.Guilford, CT: Globe Pequot.

Frank, C., Land, W. M., and Schack, T. (2013). Mental representation andlearning: the influence of practice on the development of mental represen-tation structure in complex action. Psychol. Sport Exerc. 14, 353–361. doi:10.1016/j.psychsport.2012.12.001

Gibson, J. (1977). “The theory of affordances,” in Perceiving, Acting, and Knowing,eds R. Shaw and J. Bransford (Hillsdale, NJ: Erlbaum), 67–82.

Grèzes, J., Tucker, M., Armony, J., Ellis, R., and Passingham, R. E. (2003). Objectsautomatically potentiate action: an fMRI study of implicit processing. Eur. J.Neurosci. 17, 2735–2740. doi: 10.1046/j.1460-9568.2003.02695.x

Güldenpenning, I., Koester, D., Kunde, W., Weigelt, M., and Schack, T. (2011).Motor expertise modulates the unconscious processing of human body pos-tures. Exp. Brain Res. 213, 383–391. doi: 10.1007/s00221-011-2788-7

Güldenpenning, I., Steinke, A., Koester, D., and Schack, T. (2013). Athletes andnovices are differently capable to recognize feint and non-feint actions. Exp.Brain Res. 230, 333–343. doi: 10.1007/s00221-013-3658-2

Hoffmann, J. (2003). “Anticipatory behavioral control,” in Anticipatory Behavior inAdaptive Learning Systems, eds M. Butz, O. Sigaud, and P. Gerards (Heidelberg:Springer), 44–65.

Hommel, B., Müsseler, J., Aschersleben, G., and Prinz, W. (2001). The Theory ofEvent Coding (TEC): a framework for perception and action planning. Behav.Brain Sci. 24, 849–937. doi: 10.1017/S0140525X01000103

Kiefer, M., Sim, E.-J., Liebich, S., Hauk, O., and Tanaka, J. (2007). Experience-dependent plasticity of conceptual representations in human sensory-motorareas. J. Cogn.Neurosci. 19, 525–542. doi: 10.1162/jocn.2007.19.3.525

Koch, I., Keller, P., and Prinz, W. (2004). The ideomotor approach to action control:implications for skilled performance. Int. J. Sport Exerc. Psychol. 2, 362–375. doi:10.1080/1612197X.2004.9671751

Kunde, W., Kiesel, A., and Hoffmann, J. (2003). Conscious control over thecontent of unconscious cognition. Cognition 88, 223–242. doi: 10.1016/S0010-0277(03)00023-4

Labeye, E., Oker, A., Badard, G., and Versace, R. (2008). Activation and integra-tion of motor components in a short-term priming paradigm. Acta Psychol. 129,108–111. doi: 10.1016/j.actpsy.2008.04.010

Land, W., Volchenkov, D., Bläsing, B. E., and Schack, T. (2013). From actionrepresentation to action execution: exploring the links between cognitive andbiomechanical levels of motor control. Front. Comput. Neurosci. 7, 1–25. doi:10.3389/fncom.2013.00127

Lander, H., and Lange, K. (1996). Untersuchung zur Struktur- undDimensionsanalyse begrifflich-repräsentierten Wissens. Z. Psychol. 204,55–74.

Masson, M. E., Bub, D. N., and Breuer, A. T. (2011). Priming of reach andgrasp actions by handled objects. J. Exp. Psychol. Hum. Percept. Perform. 37,1470–1484. doi: 10.1037/a0023509

Pezzulo, G., Barca, L., Bocconi, A. L., and Borghi, A. M. (2010). When affor-dances climb into your mind: advantages of motor simulation in a memorytask performed by novice and expert rock climbers. Brain Cogn. 73, 68–73. doi:10.1016/j.bandc.2010.03.002

Rand, W. M. (1971). Objective criteria for the evaluation of clustering methods.J. Am. Stat. Assoc. 66, 846–850. doi: 10.1080/01621459.1971.10482356

Santos, J. M., and Embrechts, M. (2009). “On the use of the adjusted randindex as a metric for evaluating supervised classification,” in Artificial NeuralNetworks–ICANN 2009, eds C. Alippi, M. Polycarpou, C. Panayiotou, and G.Ellinas (Berlin; Heidelberg: Springer-Verlag), 175–184. doi: 10.1007/978-3-642-04277-5_18

Schack, T. (2004). The cognitive architecture of complex movement. Int. J. SportExerc. Psychol. 2, 403–438. doi: 10.1080/1612197X.2004.9671753

Schack, T. (2012). “Measuring Mental Representations,” in Measurement in Sportand Exercise Psychology, eds G. Tenenbaum, R. C. Eklund, and A. Kamata(Champaign, IL: Human Kinetics), 203–214.

Schack, T., and Mechsner, F. (2006). Representation of motor skills in human long-term memory. Neurosci. Lett. 391, 77–81. doi: 10.1016/j.neulet.2005.10.009

Schack, T., and Ritter, H. (2009). The cognitive nature of action—functional linksbetween cognitive psychology, movement science, and robotics. Prog. Brain Res.174, 231–250. doi: 10.1016/S0079-6123(09)01319-3

Spiegel, M. A., Koester, D., and Schack, T. (2013). The functional role ofworking memory in the (re-) planning and execution of grasping move-ments. J. Exp. Psychol. Hum. Pecept. Perform. 39, 1326–1339. doi: 10.1037/a0031398

Stock, A., and Stock, C. (2004). A short history of ideo-motor action. Psychol. Res.68, 176–188. doi: 10.1007/s00426-003-0154-5

Stöckel, T., Hughes, C. M., and Schack, T. (2012). Representation of grasp pos-tures and anticipatory motor planning in children. Psychol. Res. 76, 768–776.doi: 10.1007/s00426-011-0387-7

Tucker, M., and Ellis, R. (2001). The potentiation of grasp types during visual objectcategorization. Vis. Cogn. 8, 769–800. doi: 10.1080/13506280042000144

Tucker, M., and Ellis, R. (2004). Action priming by briefly presented objects. ActaPsychol. 116, 185–203. doi: 10.1016/j.actpsy.2004.01.004

Weigelt, M., Ahlmeyer, T., Lex, H., and Schack, T. (2011). The cognitive representa-tion of a throwing technique in judo experts - Technological ways for individualskill diagnostics in high-performance sports. Psychol. Sport Exerc. 12, 231–235.doi: 10.1016/j.psychsport.2010.11.001

Weisberg, J., van Turennout, M., and Martin, A. (2007). A neural system forlearning about object function. Cereb. Cortex 17, 513–521. doi: 10.1093/cer-cor/bhj176

World Medical Assocition. (2008). Declaration of Helsinki: Ethical Principles forMedical Research Involving Human Subjects. 6th revision.


Received: 08 May 2014; paper pending published: 15 June 2014; accepted: 24 August2014; published online: 15 September 2014.Citation: Bläsing BE, Güldenpenning I, Koester D and Schack T (2014) Expertiseaffects representation structure and categorical activation of grasp postures in climbing.Front. Psychol. 5:1008. doi: 10.3389/fpsyg.2014.01008This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Bläsing, Güldenpenning, Koester and Schack. This is an open-access article distributed under the terms of the Creative Commons Attribution License(CC BY). The use, distribution or reproduction in other forums is permitted, providedthe original author(s) or licensor are credited and that the original publication in thisjournal is cited, in accordance with accepted academic practice. No use, distribution orreproduction is permitted which does not comply with these terms.












ORIGINAL RESEARCH ARTICLEpublished: 23 December 2014doi: 10.3389/fpsyg.2014.01482

Timing skills and expertise: discrete and continuous timedmovements among musicians and athletesThenille Braun Janzen1,2, William Forde Thompson1*, Paolo Ammirante3 and Ronald Ranvaud2

1 Department of Psychology, Macquarie University, Sydney, NSW, Australia2 Department of Neuroscience and Behavior, Institute of Psychology, University of São Paulo, São Paulo, Brazil3 Department of Psychology, Ryerson University, Toronto, ON, Canada

Edited by:


Reviewed by:

Bettina E. Bläsing, BielefeldUniversity, GermanyLawrence Baer, ConcordiaUniversity, Canada

*Correspondence:

William Forde Thompson,Department of Psychology,Macquarie University, Building C3A,Level 5, Macquarie Drive, Sydney,NSW 2109, Australiae-mail: [email protected]

Introduction: Movement-based expertise relies on precise timing of movements and thecapacity to predict the timing of events. Music performance involves discrete rhythmicactions that adhere to regular cycles of timed events, whereas many sports involvecontinuous movements that are not timed in a cyclical manner. It has been proposedthat the precision of discrete movements relies on event timing (clock mechanism),whereas continuous movements are controlled by emergent timing. We examinedwhether movement-based expertise influences the timing mode adopted to maintainprecise rhythmic actions.

Materials and Method: Timing precision was evaluated in musicians, athletes and controlparticipants. Discrete and continuous movements were assessed using finger-tapping andcircle-drawing tasks, respectively, based on the synchronization-continuation paradigm. InExperiment 1, no auditory feedback was provided in the continuation phase of the trials,whereas in Experiment 2 every action triggered a feedback tone.

Results: Analysis of precision in the continuation phase indicated that athletes performedsignificantly better than musicians and controls in the circle-drawing task, whereasmusicians were more precise than controls in the finger tapping task. Interestingly,musicians were also more precise than controls in the circle-drawing task. Results alsoshowed that the timing mode adopted was dependent on expertise and the presence ofauditory feedback.

Discussion: Results showed that movement-based expertise is associated with enhancedtiming, but these effects depend on the nature of the training. Expertise was foundto influence the timing strategy adopted to maintain precise rhythmic movements,suggesting that event and emergent timing mechanisms are not strictly tied to specifictasks, but can both be adopted to achieve precise timing.

Keywords: emergent timing, event timing, expertise, training, music, sports

INTRODUCTIONExperts such as musicians and athletes rely on precise timing ofbodily movements. However, whereas musicians are especiallyskilled at discrete rhythmic actions that adhere to regular cyclesof timed events (meter and pulse) (Repp and Doggett, 2007; Baeret al., 2013; Albrecht et al., 2014), athletic sports often involvefluid and continuous movements that are not timed in a cycli-cal manner (Sternad et al., 2000; Jaitner et al., 2001; Jantzen et al.,2008; Balague et al., 2013). Research suggests that the timing ofdiscrete movements (i.e., those preceded and followed by a periodwithout motion) and continuous movements depend on differ-ent strategies or processes (Robertson et al., 1999; Zelaznik et al.,2002; Huys et al., 2008; Zelaznik and Rosenbaum, 2010; Studenkaet al., 2012). Specifically, the timing of discrete movements isthought to involve a clock-like mechanism that incorporates anexplicit representation of the time interval delineated by eachdiscrete movement. In contrast, activities that involve smooth

and continuous rhythmic movements are thought to be basedon emergent timing, whereby timing regularity emerges in theabsence of a representation of time interval from the control ofparameters such as movement trajectory and velocity.

The hypothesis that event and emergent timing are distinctand dissociable systems is supported by a substantial body of evi-dence. Behavioral studies have shown that temporal variability infinger tapping is usually uncorrelated with variability in continu-ous circle drawing (Robertson et al., 1999; Zelaznik et al., 2005),and that event-timed movements, such as tapping, are signifi-cantly more precise and adjust faster to timing perturbations thancontinuous movements such as circle drawing (Elliot et al., 2009;Repp and Steinman, 2010; Studenka and Zelaznik, 2011). There isalso neurological (Ivry et al., 2002; Spencer et al., 2003, 2005) andneuroimaging (Schaal et al., 2004; Spencer et al., 2007) evidencethat event and emergent timing processes recruit different brainareas.

www.frontiersin.org December 2014 | Volume 5 | Article 1482 | 86














Braun Janzen et al. Timing skills and expertise

However, recent results have raised doubts that discrete andcontinuous movements always engage event and emergent tim-ing mechanisms, respectively (Jantzen et al., 2002, 2004; Reppand Steinman, 2010; Studenka et al., 2012; Studenka, 2014).For example, evidence suggests that the presence of perceptualevents marking the completion of time intervals can induceevent timing even for tasks performed with continuous move-ments (Zelaznik and Rosenbaum, 2010; Studenka et al., 2012).Computational simulations and behavioral studies also suggestthat task tempo and movement speed constraints (Huys et al.,2008; Zelaznik and Rosenbaum, 2010), as well as task order andpractice (Jantzen et al., 2002, 2004), are important influenceson the timing mechanism adopted for a certain task. Based onthe suggestion that the timing mechanisms recruited to performrhythmic movements are significantly influenced by several fac-tors, the present investigation tested whether two distinct forms ofexpertise and training (music and sport) differentially influencethe strategy that is engaged to perform movement-based timingtasks.

Practice is generally regarded in the motor learning literatureas one of the most essential predictors of motor skill acquisition(Schmidt and Lee, 1988; Smith, 2003; Tenenbaum and Eklund,2007; but see Mosing et al., 2014) and researchers have suggestedthat the amount of deliberate practice is directly associated withthe level of expertise acquired by athletes and musicians (Ericssonet al., 1993; Ericsson, 1996; Howe et al., 1998). It is well knownthat highly trained musicians are exceptionally precise in discrete-timing tasks, such as finger tapping with an auditory metronome(Repp, 2005, 2010; Repp and Doggett, 2007; Baer et al., 2013).Musicians tend to show smaller asynchronies and lower tappingvariability when tapping to a metronome compared with non-musician counterparts (Aschersleben, 2002; Repp, 2010). Musicalexpertise also seems to enhance the internal representation oftime as suggested by perceptual studies showing that training canimprove interval discrimination and perceptual sensitivity to tim-ing perturbations (Buonomano and Karmakar, 2002; Ivry andSchlerf, 2008; Repp, 2010). Research also demonstrates that musi-cianship specifically interacts with tasks associated with discretemovements, and not continuous movements (Baer et al., 2013),which is consistent with the view that emergent and event timingare distinct mechanisms (Zelaznik et al., 2000, 2005) and suggeststhat music expertise is predominantly an event-based skill (Repp,2005; Baer et al., 2013).

On the other hand, we know very little about how expertiseand training might influence the operation of emergent timingmechanisms, and whether the effect of training in one movement-based expertise might transfer to other timing skills. The timingof continuous rhythmic movements, such as leg movement dur-ing cycling, walking and running, or arm movements duringswimming or rowing, is thought to rely on emergent timingmechanisms (Kelso et al., 1981; Sternad et al., 2000; Jaitner et al.,2001; Jantzen et al., 2008; Elliot et al., 2009; Balague et al., 2013).This class of rhythmic movements is typically observed in sportactivities such as rowing, swimming, running, and cycling, andcould therefore be used as a model to study the effect of trainingin the production of precise continuous rhythmic movements.The purpose of the present study was to compare the ability of

movement-based experts from different domains to engage indiscrete and continuous movement tasks. Based on the hypothesisthat musical performance involves predominantly discrete rhyth-mic actions that rely on event timing, and that athletic sports gen-erally recruit fluid and continuous rhythmic movements basedon emergent timing, we examined whether movement-basedexpertise is associated with specific or general timing skills. Ifthe event and emergent timing processes are dependent on thenature of expertise and training, then athletes should be moreprecise in a timing task that involves continuous movementswhereas musicians should be more precise in a timing task thatinvolves discrete movements. In contrast, if musicians and ath-letes do not differ in their performance in both tasks, then thiswould suggest that timed movements are accomplished similarlyin these two groups of movement-based experts and, therefore,that event and emergent timing mechanisms are not strictly tiedto specific tasks, but may both be adopted to achieve precisetiming.

Experiment 1 compared the performance of elite athletes,highly trained musicians, and controls on finger-tapping andcircle-drawing tasks. The variability of inter-response inter-vals was measured in a synchronization-continuation paradigm.Participants were instructed to synchronize their movements witha metronome and continue the action at the same rate estab-lished by the metronome even when the pacing signal stopped(continuation phase). In Experiment 2, auditory feedback waspresented in the continuation phase in order to assess the effectof the presence of salient perceptual events on the timing mech-anism adopted to complete the tasks. Based on past research(Zelaznik and Rosenbaum, 2010; Studenka et al., 2012; Baer et al.,2013), we predicted that the presence of auditory feedback wouldinduce an event-timing strategy in the continuous movementtask, regardless of the expertise of the participants.

EXPERIMENT 1MATERIALS AND METHODSParticipantsFifteen athletes were recruited through the Macquarie UniversityElite Athlete Scholarship Program. Athletes (8 females, 7 males)were on average 21.31 years old (SD = 2.33, range 18–26 years)and had been involved in athletic training for an average of7.31 years (SD = 3.45). All athletes involved in the projectwere actively engaged in training and competing at State and/orNational level in athletic sports, such as swimming, rowing, mar-tial arts, rugby and others. None of the athletes had completedmore than 2 years of musical training or were involved in anymusical activities. Musicians (n = 13, 4 females) were recruitedthrough the Departments of Music and Psychology at MacquarieUniversity and local conservatories and universities. The aver-age age of musicians was 21.38 years (SD = 3.20, range 18–28years) and all participants had been involved in formal musictraining for at least 10 consecutive years (M = 10.85, SD = 2.38).Musicians played a range of instruments, including piano, gui-tar, and violin. Control participants (n = 17, 10 females) wereon average 21.76 years old (SD = 3.31, range 18–31 years). Noneof the participants in the control group reported any formalathletic or music training. Groups did not differ significantly

Frontiers in Psychology | Cognition December 2014 | Volume 5 | Article 1482 | 87





in mean age, F(2, 42) = 0.07, p = 0.93. All participants reportedthat they were right-handed and had no hearing or motorimpairment. Psychology undergraduate students were compen-sated with course credit, and all other participants receivedfinancial compensation for their participation. All participantsprovided informed consent and were debriefed about the goalsof the experiment.

Materials, stimuli, and procedureStimulus presentation and data collection were done using aMacbook Pro 9.2 laptop running custom software written inPython and tasks were completed using an Apple single-buttonmouse. The task widely used to induce event timing is fingertapping, whereas circle drawing is thought to typify emergenttiming (Repp and Steinman, 2010). The paradigm adopted forboth tasks was synchronization-continuation (Stevens, 1886). Foreach trial, participants first synchronized their movements (cir-cle drawing or finger tapping) with an isochronous metronomeclick for 18 clicks. The signal tones were 40 ms square waves clicksof 480 Hz presented at 74 dB. After the synchronization phase,the metronome stopped and participants were instructed to con-tinue to produce 36 more movements at the tempo set by themetronome. Within each trial, one of two metronome tempi wasused: slow (800 ms IOI) or fast (600 ms IOI).

In the finger-tapping task, participants repeatedly tapped onthe mouse with their right index finger at the tempo set by themetronome pacing signals and continued to tap at the samerate when the signal was removed. Participants heard the pacingsignals through Sennheiser HD 515 headphones with noise can-celing and reduction, which prevented participants from hearingany sound produced by the finger tap. No auditory feedback wasprovided.

In the circle-drawing task, participants repeatedly moved thecomputer mouse with the right hand in a circle in time withthe metronome and in a clockwise direction, and continued thismotion in the absence of the external timing cue. Participantstraced an unfilled circle template of 5 cm in diameter displayedon the screen with the mouse cursor, and were instructed to syn-chronize every time the path of the cursor crossed an intersectionat 270◦ of the circle with the metronome. Participants were toldthat timing precision was more relevant than drawing accuracy,and they were free to draw a circle at their preferred size.

Participants had 5 practice trials at 600 ms IOI before eachexperimental block. Trials were blocked by task, with tapping per-formed before circle drawing (Zelaznik and Rosenbaum, 2010;Studenka et al., 2012). For each task, trials were blocked by tempo,with the order of the two tempo conditions and the 10 trialswithin each tempo condition randomized independently for eachparticipant. Participants were permitted to take breaks in betweentrials at any time. The experiment took approximately 50 min.

Data analysisOnly responses in the continuation phase were analyzed as thesynchronization phase was used only to establish the pacing. Inorder to allow for acceleration commonly observed in the tran-sition from the synchronization to continuation phase (Flach,2005), only the final 30 movements were analyzed. For the

finger-tapping task, inter-response interval (IRI) was defined asthe elapsed time between sequential taps (in milliseconds) andfor the circle-drawing task, IRI was defined as the elapsed timebetween successive passes through the intersection. Outlier IRIswere identified as those 60% longer or shorter than the target IRIfor a given trial (4% of all IRIs analyzed in Experiment 1; 2% inExperiment 2), and were deleted.

Several timing measures were used. First, mean IRI within atrial served as a measure of timing accuracy. Second, to measuretiming precision we analyzed participants coefficient of varia-tion (CV), which was defined as the standard deviation of IRIswithin a trial divided by its mean IRI (SD/Mean). Lower CVscores indicate greater precision. CV can be considered a measureof total IRI variability, including slow drift in IRI over the courseof a trial, timing error, and motor implementation error. Third,dependencies between successive IRIs in each trial were measuredusing lag-one autocorrelation. Data were first linearly detrendedto remove the impact of slower drift over the course of a trialon dependencies between successive IRIs. In general, discrete-timing tasks are associated with negative lag-one autocorrelation.This has been proposed to arise from random delays in motorimplementation (Wing and Kristofferson, 1973) that occur inde-pendently of a central clock mechanism. One such delay shouldboth lengthen the IRI that it completes and shorten the one thatit initiates; the accumulation of these delays over the course ofa trial should be reflected in negative lag-one autocorrelation.Continuous-timing tasks, on the other hand, which are thoughtnot to involve a central clock mechanism, have been shownto result in non-negative lag-one autocorrelation (Zelaznik andRosenbaum, 2010; Baer et al., 2013). Thus, lag-one autocorrela-tion can serve as an index of event and emergent timing strategies.CV and lag-one autocorrelation values were averaged by task andtempo for each participant.

Finally, we sought to estimate clock and motor contributionsto timing variance (Wing and Kristofferson, 1973) using slopeanalysis (Ivry and Hazeltine, 1995). Slope analysis takes advantageof the well-established finding that timing variance increases lin-early as a function of squared target duration. Under the assump-tion that motor production is invariant across target durations,a positive slope (i.e., an increase in variance with target dura-tion) is thought to be influenced entirely by duration-dependentvariability (Studenka and Zelaznik, 2008). The intercept of thisregression line, on the other hand, is thought to be duration-independent, i.e., reflecting variability in the motor aspect of thetask (Studenka and Zelaznik, 2008). Different event-like taskshave been shown to exhibit equal slope values (Ivry and Hazeltine,1995; Green et al., 1999), suggesting a common underlying centralclock mechanism. On the other hand, (emergent) circle-drawingand (event) finger-tapping tasks have been shown to exhibit sig-nificantly different slopes (Robertson et al., 1999), suggestingdifferent timing mechanisms. Individual differences in slope arealso observed within tasks (Spencer et al., 2005; Baer et al., 2013),with lower slope values indicating less duration-dependent vari-ability. In the current study, for each participant and for each task,slope and intercept values were obtained from a linear regres-sion of detrended variance (averaged across trials) against squaredtarget durations (600 and 800 ms2).





RESULTSPreliminary analysis of mean IRI during the continuation phaserevealed that participants were accurate in maintaining the targettempi [fast tempo (600 ms IOI): M = 606; SD = 35; slow tempo(800 ms IOI): M = 818; SD = 55]. There were no significantdifferences between groups or group interactions.

Coefficient of variation (CV) scores were entered into a mixeddesign ANOVA with Task (circle drawing, finger tapping) andTempo (fast, slow) as within-subject factors and Group (athletes,musicians, controls) as between-subject factors. There was a sig-nificant main effect of Task, F(1, 42) = 251.01, p < 0.001, demon-strating that participants were more precise in the finger-tappingtask (M = 0.07) than the circle-drawing task (M = 0.23). It wasalso verified that there was no statistical difference in CV betweenthe fast and slow tempi conditions, F(1, 42) = 1.16, p = 0.28, andno significant interaction between Task and Tempo, F(1, 42) =2.25, p = 0.14.

Between-subjects analysis revealed a significant main effectof Group, F(2, 42) = 18.42, p < 0.001, and a significant inter-action between Group and Task, F(2, 42) = 16.48, p < 0.001.Independent sample t-tests revealed that on the circle-drawingtask athletes were significantly more precise than musicians,t(26) = 2.19, p = 0.03, and controls, t(30) = 7.00, p < 0.001.Musicians were significantly more precise than controls onthe circle-drawing task, t(28) = 3.37, p = 0.002. On the finger-tapping task, musicians were significantly more precise than con-trols, t(28) = 2.23, p = 0.03, while athletes were not significantlymore precise than controls, t(30) = 1.87, p = 0.07 (Figure 1). Theperformance of musicians and athletes was not significantly dif-ferent, t(26) = 0.61, p = 0.54. We also analyzed the correlation inCV between tasks for each of the groups tested. Results indicatedthat the variability in the finger-tapping task was not significantlycorrelated with the variability in the circle-drawing task for anygroup: musicians (p = 0.55), athletes (p = 0.08), and controls(p = 0.11).

Slope analysis was next performed to determine whether groupdifferences could be attributed to duration-dependent and/or

FIGURE 1 | Coefficient of Variation (CV) for the circle-drawing and

finger-tapping tasks per group in Experiment 1. Standard error bars areshown. Significant pairwise differences are marked with an asterisk(∗p < 0.05; ∗∗p < 0.001).

duration-independent sources. Although the slope analysis wasperformed with just two target tempi, slope values were almostentirely positive (44 of 46 participants in circle drawing; 45 of46 participants in finger tapping), indicating greater variance forlonger durations (slower tempo), which is consistent with themodel’s assumptions. An ANOVA on slope values revealed maineffects of Task, F(1, 42) = 21.01, p < 0.001, and Group, F(2, 42) =8.70, p < 0.001, as well as a marginal Group × Task interac-tion, F(2, 42) = 2.96, p = 0.06. As shown in Figure 2, slope valuesclosely mirrored those for CV. On the circle-drawing task, slopevalues for athletes (M = 0.009) and musicians (0.008) were sig-nificantly lower than for controls (M = 0.02, p = 0.02). However,although the trend was in the same direction, unlike the CV val-ues, athletes’ and musicians’ slope values did not differ from eachother, p = 0.88. On the finger-tapping task, slope values weresignificantly lower for musicians (M = 0.002) than for athletes(M = 0.004, p = 0.03) or controls (M = 0.006, p = 0.006); aswith the CV values, slope values for athletes and controls did notdiffer from each other (p = 0.33). Results of the correlation anal-ysis on the slope values indicated no significant intra-individualcorrelations for any group. An ANOVA on the intercept valuesrevealed no significant between-subjects effects or interactions.Taken together, the slope analysis indicates that group differenceswere duration-dependent, suggesting that they can be attributedto the functioning of a timing mechanism rather than to themotor constraints of the tasks.

One generally accepted indicator of the timing strategyadopted in a given task is found through the analysis of lag-one autocorrelation. Tasks that involve an event timing strat-egy exhibit lag-one autocorrelation values between −0.5 and 0,whereas tasks that involve emergent timing strategies are associ-ated with a non-negative lag-one autocorrelation (Zelaznik andRosenbaum, 2010; Delignieres and Torres, 2011). Our data wereonly partially consistent with expectations. One sample t-tests[with p-value set at 0.01 to control for Type I error (Zelaznik

FIGURE 2 | Slope for the circle-drawing and finger-tapping tasks per

group in Experiment 1. For each participant and for each task, slope valueswere obtained from a linear regression of detrended variance (averagedacross trials) against squared target durations (600 and 800 ms2). Lowerslope values indicate lower duration-dependent variability. Standard errorbars are shown. Significant pairwise differences are marked with anasterisk (∗p < 0.05).






and Rosenbaum, 2010; Baer et al., 2013)] showed that groupvalues were significantly negative in all conditions, which con-trasts with the expectation of non-negative lag-one autocorrela-tions in the (emergent) circle-drawing task. A repeated measuresANOVA revealed that, as expected, lag-one autocorrelation valueswere significantly more negative in the finger tapping condition(M = −0.14) than the circle-drawing condition (M = −0.11),F(1, 42) = 8.43, p < 0.001. Lag-one autocorrelation values, how-ever, did not significantly differ between fast and slow tempoconditions, F(1, 42) = 0.19, p = 0.66, and the interaction betweenTask and Tempo also did not reach statistical significance(p = 0.24).

Comparing lag-one autocorrelation scores between the dif-ferent groups, the analysis indicated that there was a sig-nificant Group × Task interaction, F(2, 42) = 6.81, p < 0.001.Examination of the percentage of individuals in each groupand condition with significantly negative lag-one correlations, asassessed by one sample t-tests on each individual’s data, revealedthat 60% of athletes and 59% of controls adopted an event-timingstrategy to perform the circle-drawing task, whereas 93% ofathletes and 88% of the controls used event timing in the finger-tapping task. That is, the percentage of athletes and controls thattended to rely on an event-timing strategy was significantly larger(p < 0.001) for the finger-tapping task than for the circle-drawingtask. Interestingly, lag-one autocorrelation values for musicianswere not significantly different between tasks (p = 0.37), andmusicians tended to adopt an event-timing strategy to performboth finger-tapping (63%) and circle-drawing tasks (85%).

DISCUSSIONThe results of Experiment 1 demonstrated that movement-basedexperts were significantly more precise than controls on bothtiming tasks. Athletes were significantly more precise than con-trols in the circle-drawing task, and musicians were more precisethan controls in the finger-tapping task (Repp, 2005; Repp andDoggett, 2007; Baer et al., 2013). This result suggests that exper-tise leads to enhanced timing precision in domain-related timingtasks and reinforces a dominant timing skill. This suggestion issupported by results showing that, whereas musicians were sig-nificantly more precise than controls in the finger-tapping task,the performance of elite athletes did not differ significantly fromcontrols. This result indicates that the group differences observedin this study can be attributed specifically to the functioning of atiming mechanism rather than motor control in general.

A novel finding of this research is that music training wasassociated with enhanced precision on a continuous-movementtask. Past research has suggested that formal music training onlyenhances precision of discrete movements but not continuousmovements (Baer et al., 2013). It should be acknowledged thatthe use of a computer mouse to perform the tasks might haveinfluenced the results. However, task constraints cannot read-ily account for the discrepancy between our results and thosereported by Baer et al. (2013). The slope analysis suggests that ourresults are best explained by the functioning of a timing mech-anism rather than by the constraints of the tasks. Hence, it canbe speculated that group differences may contribute to the dis-crepancy of results in these studies, such as number of years of

formal music training, instrument of expertise, amount of cur-rent involvement in musical activities, or age of commencementof training. Research is needed to assess the extent to which thesefactors contribute to the development of timing skills.

The finding that music training was associated with enhancedprecision on the continuous-movement task is compatible withthe hypothesis that the distinction between event and emergenttiming is not as rigid as initially proposed, and that these mech-anisms are not strictly tied to specific tasks such as tapping andcircle drawing (Jantzen et al., 2002, 2004; Repp and Steinman,2010; Studenka et al., 2012; Studenka, 2014). The hypothesisthat the dissociation between event and emergent timing is notan all-or-nothing process (Repp and Steinman, 2010; Studenkaet al., 2012) implies that the circumstances in which the differ-ent timing modes are employed are open for investigation. InExperiment 1, lag-one autocorrelation values were significantlynegative in all conditions, suggesting that participants tended toadopt an event-timing strategy for both discrete and continu-ous tasks. More specifically, approximately 60% of participantsin each group tended to adopt event-timing strategies to performthe circle-drawing task. Interestingly, whereas the percentage ofathletes and controls that adopted event timing was higher forfinger tapping than for circle drawing, the percentage of musi-cians that relied on event timing was not statistically differentbetween tasks. One interpretation of this result is that years of for-mal music training prompted participants to rely on event-timingmechanisms to perform any timed movement, even when thosemovements are continuous (Studenka et al., 2012; Baer et al.,2013).

In Experiment 2, we further explored the hypothesis thatmovement-based expertise is associated with enhanced skill indiscrete and continuous movement, while reinforcing one pre-dominant timing mode. We also reexamined recent evidence thatwhen participants are engaged in a timing task, the presence ofsalient feedback that defines the completion of cyclical time inter-vals elicits timing behavior consistent with event timing, evenfor continuous-movement tasks (Zelaznik and Rosenbaum, 2010;Studenka et al., 2012). Studenka et al. (2012) showed that theintroduction of discrete tactile events presented at the comple-tion of each cycle of movement induced event timing in a typicallyemergent timing task. This finding corroborated a previous studythat suggested that event timing can be elicited by the insertionof regular cycles of auditory feedback (Zelaznik and Rosenbaum,2010). To examine these issues, in Experiment 2 we tested whetherthe presence of auditory feedback elicits an event-timing strategyfor a circle-drawing task among participants with intense musicalor athletic training.

EXPERIMENT 2MATERIALS AND METHODSParticipantsThirty-one elite athletes (10 females) were recruited fromMacquarie University through the Elite Athlete ScholarshipProgram. Athletes’ average age was 21.06 years old (SD = 3.69,range 18–32 years) and they had been involved in athletic train-ing for an average of 8.31 years (SD = 5.55). Athletic trainingincluded sports that require discrete interactions with a ball or





other projectile (e.g., kicking, catching, or repelling a ball insoccer, rugby, or volleyball) and sports that primarily involve con-tinuous movements (e.g., strokes in swimming, cycling, rowing).None of the athletes were involved in the first experiment andnone had had more than 2 years of musical training. Musicians(n = 17, 15 females) were recruited through the Departments ofMusic and Psychology at Macquarie University and local conser-vatories and universities. The average age of musicians was 20.72years (SD = 3.52, range 18–29 years). Musicians were all cur-rently involved in music activities for a minimum of 2 h/weekand all had been involved in formal music training for at least10 consecutive years (M = 11.94, SD = 2.68). None of the par-ticipants were involved in the previous study. Musicians played arange of instruments, including piano, guitar, and violin. Controlparticipants (n = 10, 10 males) were postgraduate or profes-sional computer programmers recruited through the ComputerScience Department at Macquarie University. Participants wereon average 31.58 years old (SD = 7.21, range 22–49 years), andhad an average of 10 years of training in their area of exper-tise (SD = 6.07), and reported that they had no previous formalathletic training and no significant past or current involvementin music. Because the control group consisted of professionalsand postgraduate students, there was a significant group differ-ence in mean age [F(2, 55) = 26.09, p < 0.001]. All participantsprovided informed consent and were debriefed about the goalsof the experiment. Fifty-six participants were right-handed andtwo were left-handed, and all participants reported that they hadno hearing or motor impairment. Participants received financialcompensation for their participation.

Materials, stimuli, procedure, and data analysisStimulus presentation and data collection involved the sameequipment as in Experiment 1, with the exception that partici-pants completed the tasks using the laptop’s touch pad in orderto facilitate performance on the circle-drawing task. Proceduresand data analysis followed the protocol established in Experiment1. The main change was the introduction of auditory feedbackat the continuation phase of the task. For each trial, after par-ticipants synchronized their movements (circle drawing or fingertapping) with an isochronous metronome for 18 pacing signals,the metronome stopped and participants were instructed to con-tinue to produce 36 more movements at the tempo established bythe metronome.

For the finger-tapping task, participants repeatedly tapped onthe touch pad with their right index finger at the tempo setby the metronome. In the continuation phase, every tap trig-gered a feedback tone of 40 ms duration with a fundamentalfrequency of 480 Hz and at an intensity of 74 dB SPL. In thecircle-drawing task, participants repeatedly traced an unfilled cir-cle template of 5 cm in diameter displayed on the screen withthe mouse cursor using their right index finger in time with themetronome and continued the task in the absence of the exter-nal timing cue. Participants were told to pass the cursor overa crossing intersection at 270◦ of the circle in synchrony withthe metronome. In the continuation phase, every time the cur-sor trajectory crossed the intersection the auditory feedback wasprovided.

RESULTSParticipants were accurate in maintaining the target tempo duringthe continuation phases of trials [fast tempo (600 ms IOI): M =613 (SD = 25); slow tempo (800 ms IOI): M = 791 (SD = 39)].An analysis of mean IRI across the two tasks showed no significantgroup differences or group interactions. That is, all three groupsmaintained a similar overall tempo in the continuation phase ofthe timing tasks.

To measure timing precision, CV scores were averaged by taskand tempo for each participant and entered into a mixed designANOVA with Task (circle drawing, finger tapping) and Tempo(slow, fast) as within factors and Group (athletes, musicians,controls) as the between-subjects factor. The analysis revealeda significant main effect of Task, F(1, 55) = 4.60, p = 0.03, anda paired sample t-test confirmed that across the three groupsperformance on the finger-tapping task (M = 0.05) was signifi-cantly more precise than on the circle-drawing task (M = 0.10),t(57) = 6.87, p < 0.001. There was also a main effect of Tempo,F(1, 55) = 35.61, p < 0.001, and a significant interaction betweenTask and Tempo, F = 17.69, p < 0.001. Results indicated thatprecision was significantly better for fast tempo (M = 0.05, p <

0.001) than slow tempo (M = 0.11) in the finger-tapping task.Participants were also significantly more precise in fast tempo(M = 0.06) than slow tempo (M = 0.09, p < 0.001) in the circle-drawing task.

Between-subjects analysis indicated that there was a significantmain effect of Group, F(2, 55) = 3.23, p = 0.04, and a marginallystatistical interaction between Task, Tempo and Group, F(2, 55) =2.81, p = 0.06. Analysis of the circle-drawing task showed thatmusicians were significantly more precise than controls on thecircle-drawing task (p = 0.01), but there was no statistical dif-ference between the performance of athletes and musicians (p =0.24), or between athletes and controls (p = 0.07). A similarpattern was observed for the finger-tapping task, which alsocorroborated the results of Experiment 1: musicians were sig-nificantly more precise than controls (p = 0.04), but no othersignificant group differences were observed (athletes and controls,p = 0.14; musicians and athletes, p = 0.33; see Figure 3).

Different subgroups of athletes were included in the study(e.g., swimming, rowing, rugby, volleyball, squash, triathlon, icehockey, martial arts, and others). We also examined whether per-formance differed between athletes specializing in sports thatrequire discrete interactions with a ball or other projectile (e.g.,kicking, catching, or repelling a ball in soccer, rugby, or volleyball)and athletes trained in continuous movements (e.g., strokes inswimming, cycling, rowing). An independent sample t-test indi-cated that there was no statistical difference between athletes ofsports based on different movement class on either the circle-drawing task, t(29) = 1.40, p = 0.17, or the finger-tapping task,t(29) = 0.31, p = 0.75.

Slope analysis was next conducted to determine whether, asin Experiment 1, the group differences in CV could be isolatedto duration-dependent variability. As shown in Figure 4, a closecorrespondence was again observed. As with the CV values,only a main effect of Group was significant, F(2, 55) = 3.79,p = 0.03. Slope values, like CV values, were lower for musi-cians (M = 0.002) than athletes (M = 0.005; p = 0.03) and






FIGURE 3 | Coefficient of Variation (CV) per group in Experiment 2.

Standard error bars are shown.

FIGURE 4 | Slope for the circle-drawing and finger-tapping tasks per

group in Experiment 2. Standard error bars are shown.

controls (M = 0.005; p = 0.02), but did not differ betweenathletes and controls (p = 0.96). In contrast, an ANOVA onthe intercept values revealed no significant between-subjectseffects or interactions, and intraindividual correlations on theslope values were also not significant for any group. Thus, as inExperiment 1, group differences in total variability (as indexedby CV) could be isolated to duration-dependent variability (e.g.,arising from noise in a central timekeeping mechanism) ratherthan duration-independent differences associated with the motorimplementation of these tasks.

Previous research has suggested that the introduction ofa perceptual event, such as tactile or auditory feedback, canstrongly induce event-timing strategies (as indexed by negativelag-one autocorrelations) even for tasks performed with continu-ous, smoothly-produced movements (Zelaznik and Rosenbaum,2010; Studenka et al., 2012). Our data were generally con-sistent with these findings. One sample t-tests showed thatgroup means were significantly negative in all conditions (seeFigure 5). A mixed-design ANOVA with Task (circle drawing,finger tapping) and Tempo (slow, fast) as within-subject fac-tors, and Group (athletes, musicians, controls) as the between-subjects factor, revealed that lag-one autocorrelations values were

FIGURE 5 | Lag-one autocorrelation values averaged across tempi by

Group and Experiment on the circle-drawing and finger-tapping tasks.

Auditory feedback was provided in Experiment 2 only. Note: Groups ofparticipants in Experiment 2 are different from those in Experiment 1.

not significantly different between tasks [F(1, 55) = 0.21, p =0.64]. Lag-one autocorrelation values were significantly differentbetween fast and slow conditions in Experiment 2, F(1, 55) = 6.23,p = 0.01, and there was also a significant interaction betweenTask and Tempo (p = 0.002). Pairwise comparisons indicatedthat there was a significant difference between lag-one auto-correlation scores in the slow (M = −0.11) and fast conditions(M = −0.17) for the finger-tapping task (p = 0.001), but lag-one autocorrelation values did not significantly differ betweenthe slow (M = −0.13) and fast conditions (M = −0.12) for thecircle-drawing task (p = 0.70).

An examination of the percentage of individuals in each groupand condition with significantly negative lag-one autocorrela-tions, as assessed by one sample t-tests on each individual’s data,revealed that 90% of participants in the control group adoptedan event-timing strategy to perform the circle-drawing task inExperiment 2, and 60% of participants in this group used eventtiming to perform the finger-tapping task. Among movement-based experts, the percentage of individuals that adopted anevent-timing strategy to perform the circle-drawing and finger-tapping tasks was similar for musicians (76%) and athletes (68%).ANOVA confirmed that there was no interaction between Group(musicians, athletes, controls) and Task (finger tapping, circledrawing), F(2, 55) = 0.98, p = 0.38. Taken together, these resultssuggest that the majority of participants adopted an event-timingstrategy to perform both tasks in Experiment 2 (Table 1).

To test whether the presence of auditory feedback definingthe completion of cyclical time intervals influenced the timingstrategy adopted, we examined the percentage of individuals ineach group and condition that had significantly negative lag-one autocorrelations between Experiments 1 and 2 (see Table 1).Examination of Table 1 suggests that the use of event timingdepended on the condition and expertise of the participant. First,for the circle-drawing task, a smaller percentage of control (non-expert) participants used an event-timing strategy when therewas no auditory feedback (59%) than when there was auditoryfeedback (90%). Second, when there was no auditory feedback,musicians were more likely to use an event-timing strategy (85%)than control participants (59%).





Table 1 | Percentage of individuals with significantly negative lag-one

autocorrelation values for each group and condition in Experiment 1

(no auditory feedback) and Experiment 2 (with auditory feedback),

and Event Timing Index (ETI: the percentage of individuals with

negative lag-one autocorrelation/meanCV).

Measures of event timing

Experiment 1 Experiment 2

Circle drawing Circle drawing

Percentage (ETI) Percentage (ETI)

Musicians 85% (8.3) 76% (8.1)

Athletes 60% (5.8) 68% (6.8)

Controls 59% (3.3) 90% (8.6)

All 67% (5.2) 74% (7.5)

Finger tapping Finger tapping

Musicians 62% (13.1) 76% (16.6)

Athletes 93% (14.0) 68% (12.5)

Controls 88% (11.8) 60% (10.1)

All 82% (13.1) 70% (13.0)

However, interpreting the raw percentage of negative autocor-relation values is complicated by the fact that these values notonly reflect a tendency to adopt an event timing strategy; they alsoreflect timing variability (van Beers et al., 2013). Thus, changes inthe percentage of negative (lag one) autocorrelation values mayreflect a change in the strategies adopted by participants; a changein the average variability of timing; or both. For this reason, it isuseful to consider the percentage of negative autocorrelation val-ues relative to the average CV of participants in each condition.Thus, we defined the Event Timing Index (ETI) as the percent-age of participants with negative lag-one autocorrelations dividedby the CV averaged across these same participants. This normal-ized measure of event timing permits a meaningful comparisonbetween conditions.

Table 1 displays ETI values in parentheses. These values sug-gest that across groups and experiments, an event-timing strategywas used significantly less for circle drawing (mean ETI = 6.6)than for finger tapping (mean ETI = 13.0). For circle drawing,there was a greater tendency for control participants to adopt anevent timing strategy when there was auditory feedback (ETI =8.6) than when there was no auditory feedback (ETI = 3.3). Thisfinding suggests that, for circle drawing, the presence of a salientperceptual event defining the completion of cyclical time intervalsinfluenced the timing strategy adopted by non-experts.

Musicians, however, exhibited a comparatively strong ten-dency to employ event timing when performing the circle-drawing task regardless of whether there was or was not auditoryfeedback (ETI = 8.3 and 8.1, respectively). For finger tapping,the tendency to adopt an event-timing strategy was similar in thetwo Experiments (mean ETI in Exp 1 = 13.1; mean ETI in Exp 2= 13.0). In short, introducing a salient perceptual event demar-cating the completion of each movement cycle encouraged anevent-timing strategy for circle drawing, but not finger tapping.One interpretation of this finding is that there was already a strong

tendency to employ an event-timing strategy for finger tapping,so the inclusion of auditory feedback had no additional impacton this tendency.

DISCUSSIONThe findings of Experiment 2 confirmed that participants per-formed significantly more precisely in the finger-tapping taskthan in the circle-drawing task. The results also indicated thatprecision was significantly better for the fast-tempo condi-tion than the slow-tempo condition in both finger-tapping andcircle-drawing tasks. Previous studies have also reported signif-icant interactions between task precision and tempo, suggestingthat the timing mechanism adopted is affected by the rate oftimed movements (Huys et al., 2008; Repp, 2008; Zelaznik andRosenbaum, 2010). The slope analysis also suggests that the dif-ferences in total variability cannot be attributed to differencesassociated with the motor implementation of the tasks, but toduration-dependent variability (e.g., in event timing, a centralclock mechanism accumulates error as the interval durationincreases).

The results of Experiment 2 also confirmed that musicianswere significantly more precise than controls at both finger tap-ping and circle drawing. It can be suggested that music trainingmay engage and refine both discrete and continuous movements.One explanation for this result is that both event and emergenttiming are implicated in the accurate timing achieved by eliteperformers, and that music training leads to the enhanced oper-ation of both timing modes (Jantzen et al., 2002, 2004; Repp andSteinman, 2010; Studenka et al., 2012; Studenka, 2014). However,it is also possible that the superior performance by musicians onboth discrete and continuous tasks could be largely attributableto an enhanced central clock mechanism, as the results of the lag-one autocorrelation analysis suggested that the vast majority ofmusicians tended to employ an event-timing strategy to performboth discrete and continuous tasks.

In Experiment 2, the performance of athletes in the circle-drawing task did not differ significantly from that of controls,in contrast to the results of Experiment 1. This discrepancycould be related to the fact that 90% of controls and 68%of athletes adopted an event-timing strategy to perform thecontinuous-movement task when auditory feedback was avail-able (Experiment 2), whereas 59% of controls and 60% of athletesused event timing to perform the circle-drawing task when audi-tory feedback was not available (Experiment 1). In other words, inthe presence of auditory feedback there was a greater tendency toadopt an event timing strategy to perform the circle-drawing task,especially for participants without movement-based expertise.This finding corroborates previous evidence that the introductionof a perceptual event, such as tactile or auditory feedback, inducesan event-timing strategy even for tasks performed with continu-ous movements (Zelaznik and Rosenbaum, 2010; Studenka et al.,2012). This tendency may explain why the precision of contin-uous movements increases when auditory feedback is available(Zelaznik and Rosenbaum, 2010). More generally, it is knownthat the presence of feedback significantly enhances timing pre-cision Aschersleben et al., 2001; Aschersleben, 2002; Stennekenet al., 2006; Goebl and Palmer, 2009, and event timing is preferred






in synchronization tasks, given that discrete actions are less vari-able and quicker to adjust after perturbations in the sensory input(Elliot et al., 2009).

Taken together, the results of Experiment 2 corroborate find-ings obtained in Experiment 1 that movement-based expertisesignificantly improves timing skills, and that extensive training inmusic leads to enhanced precision for both discrete and contin-uous movements. The findings also support the hypothesis thatevent and emergent timing are not uniquely tied to specific typesof movements but can be influenced by expertise (Jantzen et al.,2002, 2004; Zelaznik and Rosenbaum, 2010), the presence of feed-back (Studenka and Zelaznik, 2011), and movement speed (Huyset al., 2008).

GENERAL DISCUSSIONThis investigation sought to examine the effects of expertise andtraining on the precision of timed movements. The results arecompatible with the view that movement-based training signif-icantly enhances the precision of timing skills, and that this effectdepends on the nature of the training. It was also observed thatexpertise is an important predictor of the timing mechanism thatis engaged during timed actions. These findings help to clarifythe distinction between event and emergent timing mechanismsby showing that expertise and training can influence the timingmode that is employed in a particular movement-based task.

Experiment 1 demonstrated that athletes were significantlymore precise in the production of continuous rhythmic move-ments, whereas musicians were significantly more precise indiscrete rhythmic movements in the absence of auditory feed-back. These results indicate that intense training and expertisecan help to improve timing precision, and corroborate the initialhypothesis that music performance relies predominantly on eventtiming (Repp and Doggett, 2007; Baer et al., 2013; Albrecht et al.,2014), whereas athletic activities tend to employ more predom-inantly smooth and continuous movements based on emergenttiming (Kelso et al., 1981; Sternad et al., 2000; Jaitner et al., 2001;Jantzen et al., 2008; Balague et al., 2013). Thus, hours of dailypractice involving a predominant type of movement (i.e., dis-crete or continuous) may reinforce one dominant timing mode.This finding is particularly relevant to the development of educa-tional and rehabilitation programs that could greatly benefit fromactivities targeting specific classes of movements.

It is important to state, however, that actions can be imple-mented in different ways (e.g., walking vs. marching) and mayoften engage multiple mechanisms simultaneously. For exam-ple, playing the piano not only requires precise timing of thepianist’ keystrokes but also a fluid transition of the hand acrossthe piano keys. Rowing or swinging a badminton racquet, onthe other hand, are continuous actions in the sense that themovement is not smooth and interrupted; however they are dis-crete insofar as movements are segmented by perceptual events(e.g., contact of the oar with the water, and the racquet withthe shuttlecock). Therefore, whereas it is possible to isolate dis-crete and continuous movements in laboratory for experimentalpurposes, performances often require both classes of rhythmicactions (Sternad et al., 2000; Hogan and Sternad, 2007; Sternad,2008; Repp and Steinman, 2010; Studenka, 2014). The results of

the present study do not support the idea that musical and ath-letic skill are associated with event-timing and emergent timing,respectively. On the contrary, our findings suggest that to accu-rately perform timing tasks at high skill level, experts may relyon both timing modes, although one timing mechanism is oftendominant. Therefore, an essential skill in movement-based exper-tise is to smoothly transition between movements of differentclasses.

Our findings are consistent with the idea that event and emer-gent timing mechanisms are not strictly tied to specific tasks(Jantzen et al., 2002, 2004; Repp and Steinman, 2010; Studenkaet al., 2012; Studenka, 2014). First, musicians were not only sig-nificantly more precise than controls in the finger-tapping taskbut also in the circle-drawing task, suggesting that music train-ing refines both discrete and continuous rhythmic movements.Second, lag-one autocorrelation values were significantly nega-tive in all conditions in Experiment 1, suggesting that participantstended to adopt an event-timing strategy to perform both discreteand continuous tasks, even when no salient perceptual event waspresent.

The analysis of the Event Timing Index (ETI) allowed us tofurther investigate the effect of auditory feedback on the tim-ing strategy adopted to perform continuous movements. Theseresults suggested that the percentage of musicians that used anevent-timing strategy to complete the circle-drawing task did notchange significantly when auditory feedback was provided at theend of each movement cycle. Years of formal music training mighthave prompted participants to rely on event-timing mechanismsto complete a continuous-movement task (Studenka et al., 2012;Baer et al., 2013; van Beers et al., 2013). These findings supportthe suggestion that expertise and training are important predic-tors of the timing mechanism engaged in maintaining precisetimed actions. On the other hand, the percentage of partici-pants who adopted event timing to complete the circle-drawingtask significantly increased when auditory feedback was present,especially for control participants. For this task, 59% of partici-pants tended to use an event-timing strategy when no auditoryfeedback was provided (Experiment 1), but 90% of the controlgroup adopted an event-timing strategy when auditory feed-back was provided (Experiment 2). This finding indicates thatsalient events (e.g., auditory, tactile) signaling the completion ofa movement cycle can be used to generate an internal represen-tation of the time intervals to be produced based on clock-likemechanisms (Zelaznik and Rosenbaum, 2010; Studenka et al.,2012). It is known that sensory feedback enhances timing accu-racy (Aschersleben et al., 2001; Rabin and Gordon, 2004; Repp,2005; Goebl and Palmer, 2009; Gray, 2009). However, it is impor-tant to note that the manipulation of auditory feedback is possiblein experimental conditions, but in real life circumstances mul-tiple sources of feedback may be used to monitor and refinethe accuracy and precision of timed actions (Aschersleben et al.,2001). Future studies are needed to examine the role of eventand emergent timing mechanisms in the control of discrete andcontinuous rhythmic movements in ecologically valid conditions.Such research would shed light on the relative importance ofthese two timing strategies for the production of accurately timedmovements in real-life circumstances.





It should be acknowledged that certain methodological fea-tures of our investigation limit the conclusions that can be drawn.First, we observed that lag-one autocorrelation values were sig-nificantly negative in all conditions in both experiments. Thisfinding suggests that participants tended to adopt an event-timingstrategy to perform both discrete and continuous tasks, evenwhen no salient perceptual event was present. One explanationfor this finding is that participants took part in the tapping taskbefore the circle-drawing task, as in previous studies (Zelaznikand Rosenbaum, 2010; Studenka et al., 2012), and task ordermay have significantly influenced the timing strategy adoptedby some participants. Previous research has suggested that prac-ticing one timing task reinforces a particular timing strategy,which may then persist over time and over tasks (Jantzen et al.,2002, 2004; Studenka et al., 2012). Interestingly, this carry-overeffect may have been stronger for some participants than others.Nonetheless, this possibility also corroborates a central conclu-sion of the study: that timing strategies are not strictly tied tospecific tasks but may be influenced by factors such as task order,expertise and training, and the presence of salient perceptualevents.

Second, it is important to acknowledge that different groupsof participants were used in the two experiments, preventing awithin-subject comparison between the results of these experi-ments. For this reason, it is difficult to estimate that the preciseeffect of auditory feedback on the timing strategy adopted. Toovercome this limitation, we developed the “Event Timing Index(ETI).” This index is essentially the relationship between the per-centage of participants with negative lag-one autocorrelationsdivided by CV averaged across these same participants. Theresults of this analysis strongly suggest that auditory feedbackinfluenced the timing strategy adopted for circle drawing butnot finger tapping. Further research is required to validate thisconclusion.

In summary, expertise in sports and music is significantlyassociated with enhanced precision of timing skills, but thiseffect depends on the nature of the expertise and the pres-ence of auditory feedback. It should be emphasized that oneinterpretation of these findings is that individuals with supe-rior timing precision gravitated to these pursuits. However, itis likely that expertise and training further helped to engageand refine mechanisms associated with skilled timing. Expertisewas also an important predictor of the type of timing mech-anism that individuals employed for both discrete and con-tinuous movements, which casts further doubt on the long-standing assumption that event and emergent timing mecha-nisms are strictly tied to discrete and continuous movement tasks,respectively.

AUTHOR CONTRIBUTIONSThenille Braun Janzen was the major contributor to this co-authored paper and took primary responsibility for recruitingparticipants, conducting the experiment, analysing and inter-preting the data and preparing the manuscript. William FordeThompson, Paolo Ammirante, and Ronald Ranvaud providedinput on one or more of the experimental design, data analysisand interpretation, and manuscript preparation.

ACKNOWLEDGMENTSThis research was supported in part by grants from the NationalCouncil of Technological and Scientific Development (CNPqBrazil) and a grant awarded to William Forde Thompson bythe Centre for Elite Performance, Expertise, and Training atMacquarie University. The authors are grateful to Glenn Warryof the Macquarie University Sport and Aquatic Centre for hisassistance in recruiting athletes, and to Alex Chilvers for researchassistance.

REFERENCESAlbrecht, S., Janssen, D., Quarz, E., Newell, K. M., and Schöllhorn, W. I. (2014).

Individuality of movements in music: finger and body movements during play-ing of the flute. Hum. Mov. Sci. 35, 131–144. doi: 10.1016/j.humov.2014.03.010

Aschersleben, G. (2002). Temporal control of movements in sensorimotor synchro-nization. Brain Cogn. 48, 66–69. doi: 10.1006/brcg.2001.1304

Aschersleben, G., Gehrke, J., and Prinz, W. (2001). Tapping with peripheral nerveblock: a role for tactile feedback in the timing of movements. Exp. Brain Res.136, 331–339. doi: 10.1007/s002210000562

Baer, L. H., Thibodeau, J. L. N., Gralnick, T. M., Li, K. Z. H., and Penhune, V. B.(2013). The role of musical training in emergent and event-based timing. Front.Hum. Neurosci. 7:191. doi: 10.3389/fnhum.2013.00191

Balague, N., Torrents, C., Hristovski, R., Davids, K., and Araújo, D. (2013).Overview of complex systems in sport. J. Syst. Sci. Complex. 26, 4–13. doi:10.1007/s11424-013-2285-0

Buonomano, D. V., and Karmakar, U. R. (2002). How do we tell time? Neuroscientist8, 42–51. doi: 10.1177/107385840200800109

Delignieres, D., and Torres, K. (2011). Event-based and emergent timing:dichotomy or continuum? A reply to Repp and Steinman (2010). J. Mot. Behav.43, 311–318. doi: 10.1080/00222895.2011.588274

Elliot, M. T., Welchman, A. E., and Wing, A. M. (2009). Being discrete helpskeep to the beat. Exp. Brain Res. 192, 731–737. doi: 10.1007/s00221-008-1646-8

Ericsson, K. A. (ed.). (1996). The Road to Excellence: The Acquisition of ExpertPerformance in the Arts And Sciences, Sports, and Games, 1st Edn. New York,NY: Lawrence Erlbaum Associates.

Ericsson, K. A., Krampe, R. T., and Tesch-Römer, C. (1993). The role of deliberatepractice in the acquisition of expert performance. Psychol. Rev. 100, 363–406.doi: 10.1037/0033-295X.100.3.363

Flach, R. (2005). The transition from synchronization to continuation tapping.Hum. Mov. Sci. 24, 465–483. doi: 10.1016/j.humov.2005.09.005

Goebl, W., and Palmer, C. (2009). Synchronization of timing andmotion among performing musicians. Music Percept. 26, 427–438. doi:10.1525/mp.2009.26.5.427

Gray, R. (2009). How do batters use visual, auditory, and tactile informationabout the success of a baseball swing? Res. Q. Exerc. Sport 80, 491–501. doi:10.1080/02701367.2009.10599587

Green, J. T., Ivry, R. B., and Woodruff-Pak, D. S. (1999). Timing in eyeblinkclassical conditioning and timed-interval tapping. Psychol. Sci. 10, 19–23. doi:10.1111/1467-9280.00100

Hogan, N., and Sternad, D. (2007). On rhythmic and discrete movements: reflec-tions, definitions and implications for motor control. Exp. Brain Res. 181, 13–30.doi: 10.1007/s00221-007-0899-y

Howe, M. J., Davidson, J. W., and Sloboda, J. A. (1998). Innate talents: reality ormyth? Behav. Brain Sci. 21, 399–407. doi: 10.1017/S0140525X9800123X

Huys, R., Studenka, B. E., Rheaume, N. L., Zelaznik, H. N., and Jirsa, V. K. (2008).Distinct timing mechanisms produce discrete and continuous movements. PLoSComput. Biol. 4:e1000061. doi: 10.1371/journal.pcbi.1000061

Ivry, R., and Hazeltine, R. E. (1995). The perception and production of tem-poral intervals across a range of durations: evidence for a common timingmechanism. J. Exp. Psychol. Hum. Percept. Perform. 21, 1–12. doi: 10.1037/0096-1523.21.1.3

Ivry, R. B., and Schlerf, J. E. (2008). Dedicated and intrinsic models of timeperception. Trends Cogn. Sci. 12, 273–280. doi: 10.1016/j.tics.2008.04.002

Ivry, R. B., Spencer, R. M., Zelaznik, H. N., and Diedrichsen, J. (2002). The cerebel-lum and event timing. Ann. N.Y. Acad. Sci. 978, 302–317. doi: 10.1111/j.1749-6632.2002.tb07576.x






Jaitner, T., Mendoza, L., and Schöllhorn, W. (2001). Analysis of the long jumptechnique in the transition from approach to takeoff based on time-continuouskinematic data. Eur. J. Sport Sci. 1, 1–12. doi: 10.1080/17461390100071506

Jantzen, K. J., Oullier, O., and Kelso, J. A. S. (2008). Neuroimaging coor-dination dynamics in the sport sciences. Methods 45, 325–335. doi:10.1016/j.ymeth.2008.06.001

Jantzen, K. J., Steinberg, F. L., and Kelso, J. A. S. (2002). Practice-dependentmodulation of neural activity during human sensoriomotor coordination: afunctional magnetic resonance imaging study. Neurosci. Lett. 332, 205–209. doi:10.1016/S0304-3940(02)00956-4

Jantzen, K. J., Steinberg, F. L., and Kelso, J. A. S. (2004). Brain networks underlyinghuman timing behavior are influenced by prior context. Proc. Natl. Acad. U.S.A.101, 6815–6820. doi: 10.1073/pnas.0401300101

Kelso, J. A. S., Holt, K. G., Rubin, P., and Kugler, P. N. (1981). Patterns ofhuman interlimb coordination emerge from the properties of non-linear, limitcycle oscillatory processes: theory and data. J. Mot. Behav. 13, 226–261. doi:10.1080/00222895.1981.10735251

Mosing, M. A., Madison, G., Pederson, N. L., Kuja-Halkola, R., and Ullén, F. (2014).Practice does not make perfect: no causal effect of music practice on musicability. Psychol. Sci. 25, 1795–1803. doi: 10.1177/0956797614541990

Rabin, E., and Gordon, A. M. (2004). Tactile feedback contributes to consis-tency of finger movements during typing. Exp. Brain Res. 155, 362–369. doi:10.1007/s00221-003-1736-6

Repp, B., and Doggett, R. (2007). Tapping to a very slow beat: a com-parison of musicians and nonmusicians. Music Percept. 24, 367–376. doi:10.1525/mp.2007.24.4.367

Repp, B. H. (2005). Sensorimotor synchronization: a review of the tapping litera-ture. Psychon. Bull. Rev. 12, 969–992. doi: 10.3758/BF03206433

Repp, B. H. (2008). Perfect phase correction in synchronization with slow auditorysequences. J. Mot. Behav. 40, 363–367. doi: 10.3200/JMBR.40.5.363-367

Repp, B. H. (2010). Sensorimotor synchronization and perception of timing:effects of music training and task experience. Hum. Mov. Sci. 29, 200–213. doi:10.1016/j.humov.2009.08.002

Repp, B., and Steinman, S. R. (2010). Simultaneous event-based and emergenttiming: synchronization, continuation, and phase correction. J. Mot. Behav. 42,111–126. doi: 10.1080/00222890903566418

Robertson, S. D., Zelaznik, H. N., Lantero, D. A., Gadacz, K. E., Spencer, R. M.,Doffin, J. G., et al. (1999). Correlations for timing consistency among tappingand drawing tasks: evidence against a single timing process for motor control.J. Exp. Psychol. Hum. Percept. Perform. 25, 1316–1330. doi: 10.1037/0096-1523.25.5.1316

Schaal, S., Sternard, D., Osu, R., and Kawato, M. (2004). Rhythmic arm movementis not discrete. Nat. Neurosci. 7, 1136–1143. doi: 10.1038/nn1322

Schmidt, R. A., and Lee, T. (1988). Motor Control and Learning: A BehavioralEmphasis, 5th Edn. Champaign, IL: Human kinetics.

Smith, D. J. (2003). A framework for understanding the training process lead-ing to elite performance. Sports Med. 33, 1103–1126. doi: 10.2165/00007256-200333150-00003

Spencer, R. M. C., Ivry, R. B., and Zelaznik, H. N. (2005). Role of the cerebellumin movements: control of timing or movement transitions? Exp. Brain Res. 161,383–396. doi: 10.1007/s00221-004-2088-6

Spencer, R. M. C., Zelaznik, H. N., Diedrichsen, J., and Ivry, R. B. (2003). Disruptedtiming of discontinuous but not continuous movements by cerebellar lesions.Science 300, 1437–1439. doi: 10.1126/science.1083661

Spencer, R. M., Verstynen, T., Brett, M., and Ivry, R. (2007). Cerebellar activa-tion during discrete and not continuous timed movements: an fMRI study.Neuroimage 36, 378–387. doi: 10.1016/j.neuroimage.2007.03.009

Stevens, L. T. (1886). On the time sense. Mind 11, 393–404.Stenneken, P., Prinz, W., Cole, J., Paillard, J., and Aschersleben, G. (2006). The effect

of sensory feedback on the timing of movements: evidence from deafferentedpatients. Brain Res. 1084, 123–131. doi: 10.1016/j.brainres.2006.02.057

Sternad, D. (2008). “Towards a unified theory of rhythmic and discrete movements:behavioral, modeling and imaging results,” in Coordination: Neural, Behavioraland Social Dynamics, eds A. Fuchs and V. K. Jirsa (Berlin: Springer BerlinHeidelberg), 105–133.

Sternad, D., Dean, W. J., and Schaal, S. (2000). Interaction of rhythmic and discretepattern generators in single-joint movements. Hum. Mov. Sci. 19, 627–664. doi:10.1016/S0167-9457(00)00028-2

Studenka, B. E. (2014). Response to period shifts in tapping and circle draw-ing: a window into event and emergent components of continuous movement.Psychol. Res. doi: 10.1007/s00426-014-0578-0. [Epub ahead of print].

Studenka, B. E., and Zelaznik, H. N. (2008). The influence of dominant versus non-dominant hand on event and emergent motor timing. Hum. Mov. Sci. 27, 29–52.doi: 10.1016/j.humov.2007.08.004

Studenka, B. E., and Zelaznik, H. N. (2011). Circle drawing does notexhibit auditory-motor synchronization. J. Mot. Behav. 43, 185–191. doi:10.1080/00222895.2011.555796

Studenka, B. E., Zelaznik, H. N., and Balasubramaniam, R. (2012). The distinctionbetween tapping and circle drawing with and without tactile feedback: an exam-ination of the sources of timing variance. Q. J. Exp. Psychol. 65, 1086–1100. doi:10.1080/17470218.2011.640404

Tenenbaum, G., and Eklund, R. C. (eds.). (2007). Handbook of Sport Psychology.New York, NY: Wiley.

van Beers, R. J., van der Meer, Y., and Veerman, R. M. (2013). What autocorrela-tion tells us about motor variability: insights from dart throwing. PLoS ONE8:e64332. doi: 10.1371/journal.pone.0064332

Wing, A. M., and Kristofferson, A. B. (1973). Response delays and the timingof discrete motor responses. Percept. Psychophys. 14, 5–12. doi: 10.3758/BF03198607

Zelaznik, H. N., and Rosenbaum, D. A. (2010). Timing processes are correlatedwhen tasks share a salient event. J. Exp. Psychol. Hum. Percept. Perform. 36,1565–1575. doi: 10.1037/a0020380

Zelaznik, H. N., Spencer, R. M., and Doffin, J. G. (2000). Temporal precision in tap-ping and circle drawing movements at preferred rates is not correlated: furtherevidence against timing as a general-purpose ability. J. Mot. Behav. 32, 193–199.doi: 10.1080/00222890009601370

Zelaznik, H. N., Spencer, R. M., and Ivry, R. B. (2002). Dissociation of explicitand implicit timing in repetitive tapping and drawing movements. J. Exp.Psychol. Hum. Percept. Perform. 28, 575–588. doi: 10.1037/0096-1523.28.3.575

Zelaznik, H. N., Spencer, R. M., Ivry, R. B., Baria, A., Bloom, M., Dolansky, L.,et al. (2005). Timing variability in circle drawing and tapping: probing therelationship between event and emergent timing. J. Mot. Behav. 37, 395–403.doi: 10.3200/JMBR.37.5.395-403


Received: 01 June 2014; accepted: 02 December 2014; published online: 23 December2014.Citation: Braun Janzen T, Thompson WF, Ammirante P and Ranvaud R (2014)Timing skills and expertise: discrete and continuous timed movements among musi-cians and athletes. Front. Psychol. 5:1482. doi: 10.3389/fpsyg.2014.01482This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Braun Janzen, Thompson, Ammirante and Ranvaud. This is anopen-access article distributed under the terms of the Creative Commons AttributionLicense (CC BY). The use, distribution or reproduction in other forums is permitted,provided the original author(s) or licensor are credited and that the original publica-tion in this journal is cited, in accordance with accepted academic practice. No use,distribution or reproduction is permitted which does not comply with these terms.













doi: 10.3389/fpsyg.2014.00541

Trait-based cue Utilization and initial skill acquisition:implications for models of the progression to expertiseMark W. Wiggins*, Sue Brouwers, Joel Davies andThomas Loveday

Centre for Elite Performance, Expertise, and Training, Macquarie University, North Ryde, NSW, Australia

Edited by:


Reviewed by:

Kylie Ann Steel, University of WesternSydney, AustraliaAlice F. Healy, University of Colorado,USA

*Correspondence:

Mark W. Wiggins, Centre for ElitePerformance, Expertise, and Training,Macquarie University, Balaclava Road,North Ryde, NSW 2109, Australiae-mail: [email protected]

The primary aim of this study was to examine the role of cue utilization in the initialacquisition of psycho-motor skills. Two experiments were undertaken, the first of whichexamined the relationship between cue utilization typologies and levels of accuracyfollowing four simulated, power-off landing trials in a light aircraft simulator. The resultsindicated that higher levels of cue utilization were associated with a greater level oflanding accuracy following training exposure. In the second study, participants’ levels ofcue utilization were assessed prior to two 15 min periods during which they practicedtake-offs and landings using a simulated unmanned aerial vehicle (UAV). Consistent withStudy 1, the outcomes of Study 2 revealed a statistically significant relationship amonglevels of cue utilization and the number of trials to criterion on the take-off task, and theproportion of successful trials during both take-off and landing. In combination, the resultssuggest that the capacity for the acquisition and the subsequent utilization of cues is animportant predictor of skill acquisition, particularly during the initial stages of the process.The implications for theory and applied practice are discussed.

Keywords: skill acquisition, cue utilization, expertise, training

INTRODUCTIONExpert performance across a range of environments, includingsport, medical diagnosis, and financial decision making, is char-acterized by rapid, accurate responses, even in highly complexsituations (Farrington-Darby and Wilson, 2006; Müller et al.,2006; Sherbino et al., 2012). Since this level of performance isgenerally acquired over extensive periods of exposure, there is anassumption that the capacity for sustained, high-levels of perfor-mance derives from the gradual development of highly specializedassociations or routines that are subsequently retained in mem-ory (Ericsson and Lehmann, 1996; Ericsson and Towne, 2010).Where these routines are available, they are activated rapidly and,in some cases, in the absence of conscious processing (Salthouse,1991; Finkbeiner and Forster, 2007).

One of the advantages associated with the availability of highlyspecialized routines is that their activation imposes relatively fewerdemands on working memory resources (Chung and Byrne, 2008).This enables experts to undertake multiple tasks simultaneouslyand with relatively consistent levels of accuracy (Houmourtzoglouet al., 1998; Boot et al., 2008). However, it also ensures that cog-nitive resources are available to enable both the acquisition ofadditional skills, and the refinement of those skills that havealready been acquired.

The necessity for cognitive resources to facilitate the acquisitionof cognitive skills reflects the theoretically important role of work-ing memory in enabling the association between environmentalfeatures and objects or events. For example, in the case of produc-tion systems, Anderson et al. (2004) proposed that a productioncan only emerge when the condition and action statements are resi-dent simultaneously in working memory. In the early stages of skillacquisition, the process of problem resolution involves the recall

of declarative knowledge from long term into working memory,thereby occupying what is a finite resource. The development ofa production or condition-action statement obviates this demandfor declarative knowledge and the faster that this process occurs,the greater the capacity to allocate the residual resources to othertasks, thereby potentially improving the rate of skill acquisition.

The proposition that efficiencies in information processingcan be gained through a parsimonious association between envi-ronmental features and events or objects is a consistent themein various models of skill acquisition, as well as explanationsof the superior performance of experts (Ericsson and Kintsch,1995). Notions of bounded rationality, automated processing, andinstance processing all presuppose tightly arranged associations toexplain the rapid recognition and response to situations that arecharacteristic of expertise (Logan, 2002; Campitelli et al., 2007;Pachur and Marinello, 2013).

Klein (2011) suggests that the value of the associations in mem-ory lies in their capacity to enable an operator to quickly classify ordiagnose a situation. This process triggers an associated responsefrom memory, and thereby facilitates a relatively rapid response.Where Anderson et al. (2004) would refer to this process as theactivation of a condition–action statement in the form of a pro-duction, Gigerenzer and Gaissmaier (2011), Klein et al. (2010),and Brunswik (1955) suggest the application of cue-based asso-ciations. Cues constitute relationships among features and eventsor objects that are resident in the environment (Wiggins, 2006).They are highly specialized and targeted, and they enable the rapidrecognition and response to particular situations.

The difficulty associated with the acquisition of associationsbetween phenomena is that their coexistence does not neces-sarily infer a causal relationship (Holyoak and Cheng, 2011).














Wiggins et al. Cue utilization and skill acquisition

A “storming” process is necessary during the acquisition of skilledperformance whereby associations are quickly tested, discarded,or revised to ensure that they are both as accurate and as parsimo-nious as possible (Wiggins, 2012). However, an inevitable part ofthis process of storming is the commission of errors, whereby inap-propriate cues may be triggered, or the cue associations themselvesmay be overly general or incorrect, thereby leading to inefficienciesor delays in the acquisition of skilled performance (Bridger andMecklinger, 2014).

At a fundamental level, the capacity to acquire and subsequentlyrevise cue-based relationships requires a cognitive strategy involv-ing the identification of salient features in the environment, theperception of associations between features and events/objects,the retention of these associations in memory and, finally, therecognition of those situations during which the application ofcues applies (Wiggins, 2014). The efficiency with which this pro-cess occurs will determine the rate of skill acquisition. However,what remains unclear is whether the capacity for the developmentand application of cues is determined by a particular context, orwhether it constitutes an underlying trait so that the rate of cueacquisition within one context predicts performance in a relateddomain.

The aim of the present study was to consider the relation-ship between the cue utilization in the context of motor vehiclehazard detection and way finding, and skill acquisition in learn-ing to land a simulated aircraft and learning to take-off andland a line-of-sight umanned aerial vehicle (UAV). If skilledpsycho-motor performance is dependent upon the availability offeature–event/objects in memory in the form of cues, the rateat which cues are acquired within a given period should pre-dict the rate of skill acquisition. Moreover, if cue acquisition isa trait, then the rate of cue utilization evident in one task shouldreflect the rate of skill acquisition in tasks that demand similarcapabilities.

Although the relationship between cue acquisition and skillacquisition has yet to be examined empirically, some evidence fora relationship can be drawn from Small et al. (2014) who wereinvestigating the relationship between cue utilization and perfor-mance during a novel, short vigilance task. Of particular interestin the context of the present study was the observation that par-ticipants’ performance differed in the rate at which they becamefamiliar with a novel representation of a domain-related task. Thisdifference in the rate of skill acquisition was not explained by yearsof industry-related experience, suggesting that the acquisition ofcues may constitute an underlying trait.

STUDY 1Study 1 was designed to examine the relationship between a com-posite measure of cue utilization in the context of motor vehiclehazard and way-finding, and performance in learning to land asimulated aircraft as close as possible to a runway target. Since cueutilization represents the outcome of the process of cue acquisi-tion, it was important to control for domain-related experience.To that extent, drivers’ years of experience were recorded, andthese data were employed as a covariate to control statistically forexposure to the domain. It was hypothesized that, controlling fordriving experience, participants who recorded greater levels of cue

utilization would land closer to the runway target following fourexposure trials.

METHODParticipantsA total of 51 university students (25 male and 26 female) wererecruited for the study. These participants comprised first- andsecond-year psychology students who each received 0.5% coursecredit for their participation. Their ages ranged from 18 to 22 years(M = 20.27, SD = 1.601). The inclusion criteria comprisedlicensed drivers who had never previously flown a flight simulator.

InstrumentsA demographic questionnaire required participants to indicatetheir age, sex, years of driving experience, weekly driving fre-quency, weekly frequency of video-game play, and their experienceoperating a flight simulator. Cue utilization was assessed using theEXPERT Intensive Skills Evaluation (EXPERTise; Wiggins et al.,2010) situation judgement test (SJT). Designed to measure per-formance on several cue-based processing and problem solvingtasks, it provides a composite assessment of domain-related cueutilization.

EXPERTise incorporates experimental tasks that have sepa-rately and collectively been associated with differences in oper-ational performance. They include a paired association task,designed to establish the availability of feature–event/object rela-tionships in the form of cues, a feature identification task, designedto assess feature priming, and a feature discrimination task,designed to test the precision of cue-based associations in memory.Each task yields a distinct but complementary assessment of cueutilization that, in combination, provides an overall assessment ofthe utilization of cues in memory.

In the paired association task, participants were presented withtwo different terms (feature–event/object) that appeared adjacentto one another for 1800 ms. Using a six-point Likert scale, partic-ipants were asked to indicate the extent to which they consideredthe two words related. Examples included related terms such as“journey time” (event) with “car speed” (feature) and relativelyless related terms such as “red traffic light” (feature) and “freeway”(object). Higher levels of cue utilization were associated with agreater variance in the perceived relatedness of terms (Ackermanand Rathburn, 1984; Morrison et al., 2013). The use of variance asa measure of cue utilization in this context is based on the assump-tion that, through experience, the associations between cues arebetter refined and thereby lead to a greater level of dichotomy inperceptions of the association between features and events/objects.The measure of performance has been used successfully to differ-entiate experts from non-experts in a range of context (Wittemanet al., 2012)

For the feature identification task, participants were tasked withlocating, as quickly as possible, a ball displayed at different loca-tions within a static, complex driving scene; a process similar tomeasures of field dependence (Goodenough, 1976). In this case,lower response latencies reflect a greater capacity to extract keyfeatures from a complex array (Wiggins, 2014).

In the feature discrimination task, participants were providedwith a hypothetical scenario (“After driving for some time, you





become aware that you have lost your bearings... Surveying yoursurrounds, you see several cars driving with surfboards on theirroofs. There are also several beachgoers and shoppers walkingnearby. . .. In the distance, you can see high-rise apartment blocksas well as palm trees. There are also several street signs visible. . .”).After making a decision as to their most likely response underthe circumstances, participants rated the importance of differentfeatures in the formulation of their response. In the driving ver-sion of EXPERTise, 14 features were presented during the featurediscrimination task, to which participants responded using a 10-point Likert Scale, with higher ratings equating to a greater levelof importance in the decision process. In this task, higher levelsof cue utilization were reflected in higher variance in the ratingsof the perceived relevance of features (Weiss and Shanteau, 2003;Witteman et al., 2012).

The construct validity of EXPERTise has been demonstratedin a number of different domains, whereby typologies that wereformed on the basis of performance across the EXPERTise tasksdifferentiated both simulated and actual performance in the work-place (Loveday et al., 2013b,c; Loveday and Wiggins, 2014; Wigginset al., 2014). The test–retest reliability of the typologies has beendemonstrated in the context of power control operators at sixmonthly intervals, κ = 0.59 (Loveday et al., 2013a).

Flight simulatorA Redbird FMX-1000 flight simulator incorporating three degreesof freedom was used to position a simulated Cessna 172 at analtitude of 1000 feet and a distance of 1.5 km from the runway.Two large white bars positioned on the runway represented thetarget landing point. After landing, the aircraft was repositionedand this process was repeated for four trials.

ProcedureThe participants completed the tasks individually, and all measureswere presented sequentially in the same sitting. Having completedthe demographic questionnaire, participants were directed to theon-line version of EXPERTise and they were asked to follow theinstructions that were displayed on the computer screen. Oncecompleted, participants entered the flight simulator and werebriefed on the basic controls and the aim of the exercise. Theythen completed four trials, attempting to guide the aircraft to alanding on the runway.

RESULTSCue utilization typologiesPrior to detailed analysis, it was necessary to identify the cue uti-lization typologies that corresponded to relatively higher and lowerlevels of cue utilization. These typologies were based on the out-comes of the EXPERTise tasks and were employed in this case dueto the correspondence with previous methodological approachesto the application of EXPERTise-related outcomes (Loveday et al.,2013b,c; Loveday and Wiggins, 2014; Wiggins et al., 2014). The cal-culation of typologies began with the aggregation of the responseswithin the tasks, the calculation of z-scores, and a cluster analysisto identify whether two, meaningful typologies could be estab-lished. In the present study, two typologies were identified withcentroids that corresponded to: (a) a lower response latency in the

feature identification task, and higher variance in the paired asso-ciation and feature discrimination tasks (higher cue utilization),and (b) a greater response latency in the feature identificationtask, and lower variance in the paired association and featurediscrimination tasks (lower cue utilization). The cluster analysisclassified 15 participants in the higher cue utilization typologyand 33 participants in the lower cue utilization typology (seeTable 1). The data for three participants were excluded due tomissing data.

Landing performanceFlight performance data comprised the primary dependent vari-able in the current study. Specifically, the flight task requiredparticipants to land the aircraft at a specific target located on therunway. The difference between a participant’s landing position(longitude and latitude) and the ideal landing location (longi-tude and latitude) formed the “flight performance” variable inthe current study. This was measured in kilometers and calcu-lated using a distance calculator for compass coordinates. Lowerscores in flight performance represent a shorter distance fromthe landing target, and thus, greater accuracy during the flighttask.

Five landing trials were conducted by each participant.The Shapiro–Wilks normality statistic for each landing trialwas non-significant (p < 0.05) and inspection of the P–Pplots indicated normal distributions. The landing trials wereanalyzed via repeated measures and post hoc pairwise com-parisons. A significant main effect for landing performance,F(4,144) = 12.83, p = 0.000, suggested that landing accu-racy differed over trials. Post hoc comparisons and inspectionof means revealed a pattern of steady and statistically signif-icant improvement in landing accuracy, until the final trial(fifth landing). Landing accuracy in the first trial (M = 0.66,SD = 0.18, SE = 0.030) significantly improved in the sec-ond (M = 0.58, SD = 0.19, SE = 0.032, p = 0.011), andperformance in the second trial significantly improved in thethird trial (M = 0.51, SD = 0.21, SE = 0.034, p = 0.000).Compared to the third trial, the fourth showed continuedimprovement, with a reduced mean distance from the tar-get (M = 0.48, SD = 0.20, SE = 0.033). However, thefourth trial was not significantly different from the preced-ing trial (p = 1.00). The final trial indicated a decrease inperformance and increased error (M = 0.49, SD = 0.26,SE = 0.043). Trial five was also not significantly differentfrom trials two, three, or four. Taken together, this patternsuggests that learning occurred, and that performance in thefinal trial may have been affected by fatigue. For this reason,

Table 1 | Cluster centroids for the EXPERTise task scores for Study 1.

Cluster 1 (n = 15) Cluster 2 (n = 33)

Feature identification –1.01 0.47

Feature discrimination 0.99 –0.45

Paired association 1.24 –0.54





the landing performance for trial four was selected as thedependent variable in subsequent analyses, since it was at thispoint that optimal performance, following learning, had beenachieved.

Cue utilization and landing performanceA univariate analysis of covariance (ANCOVA) was used to testthe relationship between cue utilization and landing performance,and comprised the cue utilization typology as the independentvariable (higher and lower), landing performance (as measuredby the distance from the target on the fourth trial) as the depen-dent variable, and years of driving experience as a covariate. Theresults revealed a statistically significant main effect for cue uti-lization typology, F(1,45) = 4.18, p = 0.047. An inspection ofthe mean landing performance of the clusters indicates that par-ticipants with higher levels of cue utilization (controlling fordriving experience) landed the aircraft closer to target (Meandistance = 0.37, SD = 0.23) than participants with lower lev-els of cue utilization (Mean distance = 0.53, SD = 0.19). Thissuggests that a relationship exists between cue acquisition andthe acquisition of skilled performance in related, psycho-motortask.

Using hierarchical regression with change statistics, the vari-ance in landing performance attributable to cluster was 11.5%and when driving experience was included in the model,the total proportion of the variance explained increased to27%. This change represents an increase of 24%, and wasstatistically significant, F(1,45) = 9.49, p = 0.005. Partialcorrelations for driving experience and cluster revealed thatthe variance uniquely attributed to cluster was 8.5% andthe variance uniquely attributed to driving experience was17.4%.

STUDY 2Study 2 was designed to extend the outcomes of Study 1 byexamining the relationship between driving-related cue acquisi-tion and the development of skilled performance in the operationof a simulated UAV. If cue acquisition represents a precursorto skilled performance, then measures of cue utilization (con-trolling for driving experience) were expected to predict thenumber of trials required to reach a predetermined level of take-off and landing performance, together with the proportion ofsuccessful take-off and landing trials. Since the acquisition ofcues is also likely to dependent upon the capacity to excludeextraneous information and thereby identify predictive feature–event/object associations, it was also anticipated that measuresof sensory processing sensitivity (SPS) and attentional controlwould account for a proportion of the variance associated withthe acquisition of skilled performance in operating the UAV.Higher levels of SPS and lower levels of attentional controlare normally associated with clinical conditions and heightenedarousal (Aron and Aron, 1997; Derryberry and Reed, 2002).Therefore, lower levels of SPS in combination with higher lev-els of attentional control and greater levels of cue acquisitionwere expected to account for a greater proportion of success-ful trials and trials to reach criterion than either variable inisolation.

METHODParticipantsA total of 50 university students participated in the study ofwhom 21 were male and 29 were female. They were recruitedfrom the Psychology Research Pool, and each received 1% coursecredit for their participation. They were aged between 18 and26 (M = 18.87, SD = 1.58), possessed a current motor vehicledriver’s license, and had no experience in remote control aircraftoperation.

InstrumentsAs in Study 1, the participants completed a demographicquestionnaire, including questions related to video game anddriving experience, and then progressed to complete the on-lineversion of EXPERTise. They were subsequently asked to com-plete Aron and Aron’s (1997) Highly Sensitive Person Scale andDerryberry and Reed’s (2002) Attentional Control Scale. The 27-item Highly Sensitive Person Scale requires participants to indicatetheir response on a seven-point Likert scale. An example item is“Are you easily overwhelmed by things like bright lights, strongsmells, coarse fabrics, or sirens close by?” Levels of sensitivityare calculated by summing the responses to the questions withhigher scores associated with higher levels of SPS. The scale hasadequate discriminant, convergent, and overall construct validity,and Cronbach’s alphas have been obtained in the range of 0.81–0.84, demonstrating adequate reliability (Jagiellowicz et al., 2010).An alpha of 0.77 was achieved in the present study.

The Attentional Control Scale is designed to measure an indi-vidual’s general capacity to focus attention on a task, to filter outdistractions, shift attention between tasks, and to flexibly controlthought (Derryberry and Reed, 2002). The 20-item scale requiredparticipants to indicate their response on a four-point Likert scale.An example item is “When concentrating, I can focus my atten-tion so that I become unaware of what’s going on in the roomaround me?” Scores are calculated by summing the responses tothe questions with higher scores associated with greater levels ofAttentional Control. The scale has adequate discriminant, con-vergent, and overall construct validity in different populations(Fajkowska and Derryberry, 2010). A Cronbach’s alpha of 0.83was achieved in the present study.

UAV simulatorReal Flight 6.0TM was used to simulate the operation of a UAV.The simulator was displayed on a 40-inch monitor with con-trol exercised using a standard remote control aircraft transmitter,incorporating two joysticks, one to control the pitch and roll ofthe aircraft and the other to control power.

For the take-off task, the UAV was located at the end of therunway, and participants were asked to accelerate the aircraft usingthe joystick and fly the aircraft down the extended center line ofthe runway through two virtual parallel lines. In the case of thiscomputer program, the failure to position the aircraft within theparallel lines would result in the destruction of the aircraft, whereinthe aircraft was repositioned at the take-off in preparation for thenext trial. Similarly, in the case of the landing, the aircraft waspositioned at altitude, a short distance from the runway alongthe extended center line. Participants were asked to advance or





retard the throttle while maintaining directional control of theaircraft, and descend between two parallel lines. Similar to thetake-off task, the failure to maintain the position of the aircraftwithin the two parallel lines would result in the destruction ofthe aircraft and a return to the landing approach position. Thenumber of trials was recorded and criterion performance was setat three consecutive successful take-offs or landings to reduce theinfluence of accidental and sporadic successes.

ProcedureOnce the initial questionnaires had been completed, participantswere asked to stand in front of the UAV display and the task wasdescribed, together with instructions concerning the control ofthe simulated aircraft. Participants were advised that they wereto try and ensure that the aircraft departed or landed within theparallel lines and that flight outside the parallel lines would resultin the destruction of the aircraft with the requirement to restartthe trial. Similarly, if the participant succeeded in completing atrial successfully, the aircraft trial would be restarted. Participantswere advised that they would be given 15 min to complete as manysuccessful trials as possible. The take-off trials always preceded thelanding trials.

RESULTSThe aim of Study 2 was to examine the differential effects of cueutilization, sensory processing sensitivity, and attentional controlon trials to criterion in learning to operate a UAV simulator. Similarto Study 1, data arising from the EXPERTise SJT were aggregatedand converted into z scores. However, unlike Study 1, typologieswere not established. Instead, a grand mean was employed as anoverall, standardized measure of cue utilization and to allow forregression analyses.

UAV performanceFour measures of UAV skill acquisition were used as dependantvariables in the present study. The number of trials to achieve cri-terion performance (three consecutive successful trials) was usedto establish the rate of skill acquisition. A descriptive analysis of thedata revealed that skewness was outside normal limits. A square-root transformation was undertaken subsequently, which reducedthe skewness to an acceptable level (<1).

The number of trials to reach criterion performance could notbe used as a measure of the rate of skill acquisition, as only 31 ofthe 50 participants were able to complete three consecutive land-ings successfully. Consequently, a new categorical variable wascalculated with two levels: those who were able to achieve land-ing criterion performance and those who were unable to achievecriterion performance.

The proportion of successful trials was included as a broadermeasure of skill acquisition that was influenced by both the rateof skill acquisition and the consistency of performance beyondthe initial achievement of the criterion. Proportions were derivedfor both the take-off and landing tasks by calculating the numberof successful trials as a proportion of the total number of trialscompleted by each participant. The mean number of total trialsacross all participants was 70.82 for the take-off task (SD = 10.312)and 76.46 for the landing task (SD = 11.38).

Modeling UAV performanceA measure of SPS was obtained by summing the responses to all 27items of the Highly Sensitive Person Scale. The Attentional Con-trol score was obtained by summing the responses for the reversescored items (1, 2, 3, 6, 7, 8, 11, 12, 15, 16, and 20) from theAttentional Control Scale. A hierarchical multiple linear regres-sion was used initially to determine the relationship among cueutilization scores, weekly videogame use, SPS, Attentional Con-trol, driving experience, and the proportion of successful take-offtrials. Entering the cue utilization score explained 19.7% of thevariance in the proportion of take-off trials and was statisticallysignificant, F(1,47) = 11.81, p < 0.01. This increased to 30.3%of the explained variance with the addition of Sensory Process-ing Sensitivity scores, a change that was statistically significant,F(1,46) = 7.15, p < 0.01. The addition of the remaining variablesfailed to increase the amount of variance explained, indicating thatthe proportion of successful take-off trials during skill acquisitionof the UAV was best predicted by a combination of a higher levelof cue utilization (β = 0.38) and lower level of sensory processingsensitivity (β = –0.33).

Consistent with the results associated with the proportionof successful take-off trials, levels of cue utilization and sen-sory processing sensitivity also provided the model of bestfit for the proportion of successful landing trials using theUAV, F(2,47) = 8.33, p < 0.01. Specifically, 26.2% of thevariance in the proportion of successful landing trials waspredicted by higher levels of cue utilization (β = 0.37) andlower levels of sensory processing sensitivity (β = –0.29).Although this is slightly lower than the variance explainedfor take-off performance, it is of particular note that nei-ther driving experience, videogame experience nor attentionalcontrol contributed significantly to the final model. As mightbe expected a strong correlation was evidence between take-off and landing performance using the UAV [r(50) = 0.69,p < 0.001].

In relation to the number of trials required to satisfy the take-off criterion, the regression model of best fit was restricted to thelevel of cue utilization, β = –0.41, F(1,42) = 8.58, p < 0.01, whichexplained 17% of the variance in performance. No other vari-ables contributed significantly to the model, including sensoryprocessing sensitivity. This suggests that, while sensory processingsensitivity may explain some of the variance associated with sus-tained performance beyond the achievement of a criterion level ofperformance, it is not predictive of the initial achievement of thiscriterion.

Since the nature of the data arising from the landing per-formance task precluded the use of linear regression, a logisticregression was employed in which the dependent variable com-prised whether or not the participant achieved the landing crite-rion. Consistent with the data for the linear regression concerningthe achievement of the take-off criterion, only cue utilizationwas retained as in the model of best fit, β = 1.68, SE = 0.584,Wald’s X2 = 8.27, Exp(B) = 5.35, p = 0.004. The results sug-gest that the odds of achieving landing criterion increased by afactor of 5.35 for each unit increased in the log concentrationof cue utilization (Hosmer–Lemeshow X2 = 8.78, Cox and SnellR2 = 0.20).





GENERAL DISCUSSIONThe overall aim of this research was to examine the role of cueutilization in the early development of skilled performance intwo related domains. Cue utilization has been established pre-viously as a characteristic of expertise (Loveday et al., 2013c), andthis suggests that the capacity to identify, acquire, and retainfeature–event/object relationships in memory may be a necessaryprecursor for the acquisition of expertise. Understanding the roleof cue acquisition at the initial stages of skill development poten-tially gives rise to a more complete model of the mechanisms thatwill facilitate the transition from novice through competence toexpertise.

Study 1 sought to establish the relationship between cueutilization and performance in learning to land a simulatedaircraft. Cue utilization was assessed using the driving-relatedversion of EXPERTise, and this enabled the formation oftypologies reflecting relatively higher or lower levels of cueutilization. The results revealed a statistically significant rela-tionship between cue utilization typology and the proximity tothe runway target following the fourth landing trial. Specifi-cally, higher levels of cue utilization, controlling for drivingexposure, were associated with a closer proximity to the land-ing target. This suggests an association between a measure ofcue utilization and performance on a novel skill acquisitiontask.

To establish the generalizability of the outcomes of Study1 and to explore additional explanations of the mechanismsof skill acquisition, Study 2 was designed to employ thesame measure of cue utilization (driving-related) but consid-ered the acquisition of skilled performance in the operationof a UAV. While conceptually similar to the operation of aflight simulator, the operator of a Line-of-Sight UAV con-trols the aircraft from the ground using a remote controldevice.

In addition to cue utilization, additional variables were incor-porated into the analysis, including video-game use, SPS, andattentional control. It was surmised that, in combination withthe capacity to associate feature–event/object relationships in theform of cues, the capacity to identify prospective features andevents would depend upon the capacity to direct attention toparticularly salient features and avoid being distracted by thosefeatures that are likely to embody little predictive capacity. Theresults revealed a strong model in which a combination of cueutilization scores and sensory processing sensitivity was most pre-dictive of both the number of trials to reach criterion on thelanding and take-off tasks, together with the proportion of suc-cessful trials. Specifically, greater levels of cue utilization and lowersensory processing sensitivity predicted 31.7% of the varianceassociated with the acquisition of take-off performance on theUAV simulator.

In combination, the results of the two studies suggest that cueutilization in one context may play a significant role in the initialacquisition of skilled performance in other, related tasks. The factthat cue utilization is also characteristic of expert performancesuggests that cues may constitute a key cognitive mechanism bywhich skill acquisition occurs, even at the earliest stages of theprocess.

THEORETICAL IMPLICATIONSAlthough there have been a number of different theoretical propo-sitions concerning the cognitive mechanisms that facilitate theprogression to expertise, including cases, instances, and produc-tions, the present study targeted behavior that was most closelyassociated with the utilization of cues. EXPERTise is designed totarget a number of aspects of cue utilization, including the capabil-ity to discern key features from a complex visual background, thecapacity to differentiate the strength of the relationship betweendifferent feature–event/object pairs, and the relative importanceof features in the context of diagnosis.

On the basis of the relatively consistent relationship betweenthe driving-related version of EXPERTise and performance onthe skill acquisition tasks, it might be concluded that the capac-ity to identify features and discern the strength of relationshipsbetween features and events/objects, constitutes a capability thatinforms the acquisition of skilled performance on both a three-dimensional tracking task in the context of flight simulation,and take-offs and landings in the context of operating a UAV.Moreover, in cases involving the acquisition of novel skills, therate of progression toward expertise may be dependent upon:(a) the extent to which key features can be identified; (b) theirassociation with events/objects established and retained in mem-ory; and (c) their accurate application during the process of skillacquisition.

Although the outcomes of present research do not necessarilydiscount the role of productions as an explanation of skill acquisi-tion, given the conceptual similarities between the two constructs,the role of cases and instances as an explanation of the processis less clear. In particular, the predictive capacity of a process thatdeconstructs tasks into distinct feature–event/object relationships,coupled with the lack of domain-related knowledge on the part ofparticipants gives rise to a cue-based explanation of psycho-motorskill acquisition, particularly at the initial stages of the process.This explanation is consistent with previous research establishingthe role of cue-based training in developing the skills of novicesin other domains (Abernethy, 1990; Wiggins and O’Hare, 2003;Markovits, 2013; Momm et al., 2013).

As a context-dependent measure of cue utilization, EXPER-Tise has differentiated performance among pediatricians, powercontrollers, and software engineers. It has also identified differ-ences in the acquisition of skilled performance in the context ofpower control. In combination with the results of the presentstudy, cue utilization, as measured by EXPERTise, appearsto both differentiate the performance of different operators,and predict the rate and the achievement of skilled perfor-mance. However, despite the relative consistency of the outcomesachieved, a number of key questions remain that are rele-vant to those models of cognitive skill acquisition that positthat the progression to expertise is based on the acquisitionand utilization of cues. First, and most important, EXPER-Tise is purported to measure cue utilization but, in the absenceof neurological evidence, that argument will remain specula-tive. Second, longitudinal studies have yet to be completedthat include competent practitioners. Much of the work thusfar has focused on the performance of novices, the transi-tion from competence to expertise, or on retrospective accounts





of skill acquisition and experience from skilled performers(e.g., Young and Salmela, 2010). Little research has been under-taken that considers the key transition from novice to competencethat, potentially, is the stage at which mental models are acquiredand tested.

APPLIED IMPLICATIONSThe practical implications of the present study include the identi-fication of those practitioners who are relatively more likely toachieve criterion performance on a three-dimensional psycho-motor tracking task, together with the rate at which this progres-sion is likely to occur. The associated benefits include a potentialreduction in the costs of training and attrition due to the selectionof candidates who are capable of acquiring particular skills moreaccurately and at a rate that reduces the investment in both timeand access to expensive simulation technologies.

In addition to the selection of candidates, the results of thepresent study also enabled the development of interventions forthose candidates who, having been selected, experience plateausin their acquisition of skilled performance. The apparent key roleof cue utilization during the initial stages of skill acquisition sug-gests that learning plateaus may be explained by the inability ofthe learner either to identify predictive features, and/or establish arelationship between predictive features and associated events orobjects. Therefore, learning plateaus may be considerably short-ened if learners can be directed toward those features that are mostappropriate in the context of the problem being confronted. Evi-dence for the potential utility of this type of approach to initiallearning can be drawn from Lagnado et al. (2006) who observedthat learning environments that are directed toward the identifica-tion of feature-outcome relationships can facilitate the acquisitionof cue-based associations that, in turn, lead to improvements inperformance. Similarly, Wulf et al. (2000) observed that improve-ments in tennis could be achieved by directing learners’ attentiontoward what they referred to as the “antecedents” and “effects”of particular types of strokes from opponents. In identifying andremediating “gaps” in cue-based processing, it becomes possibleto augment existing training initiatives, thereby maintaining anoptimal rate of skill acquisition, irrespective of the nature of thelearner.

CONCLUSIONThe aim of this paper was to establish whether a measure of cueutilization predicts the rate and the achievement of initial perfor-mance in the acquisition of two, three-dimensional tracking tasks.In Study 1, participants learnt to maneuver a simulated light air-craft to land nearest to a target on a runway. Study 2 involvedlearning to take-off and land a UAV using a remote control device.In both studies, a relationship was established between cue utiliza-tion and task-related performance, whereby relatively higher levelsof cue utilization were associated with both a greater rate of skillacquisition and a greater proportion of successful trials in learningto operate the UAV. Given that cue utilization also differentiatesgreater from lesser performance among experienced operators, theresults suggest that the acquisition and subsequent utilization ofcues may play a significant role in facilitating the rate and theachievement of expertise.

REFERENCESAbernethy, B. (1990). Anticipation in squash: differences in advance cue uti-

lization between expert and novice players. J. Sports Sci. 8, 17–34. doi:10.1080/02640419008732128

Ackerman, B. P., and Rathburn, J. (1984). The effect of recognition experi-ence on cued recall in children and adults. Child Dev. 55, 1855–1864. doi:10.2307/1129932

Anderson, J. R., Bothell, D., Byrne, M. D., Douglass, S., Lebiere, C., and Qin,Y. (2004). An integrated theory of the mind. Psychol. Rev. 111, 1036. doi:10.1037/0033-295X.111.4.1036

Aron, E., and Aron, A. (1997). Sensory-processing sensitivity and its relationto introversion and emotionality. J. Pers. Soc. Psychol. 73, 345–368. doi:10.1037/0022-3514.73.2.345

Boot, W. R., Kramer, A. F., Simons, D. J., Fabiani, M., and Gratton, G. (2008). Theeffects of video game playing on attention, memory, and executive control. ActaPsychol. 129, 387–398. doi: 10.1016/j.actpsy.2008.09.005

Bridger, E. K., and Mecklinger, A. (2014). Errorful and errorless learning: the impactof cue–target constraint in learning from errors. Mem. Cogn. doi: 10.3758/s13421-014-0408-z [Epub ahead of print].

Brunswik, E. (1955). Representative design and probabilistic theory in a functionalpsychology. Psychol. Rev. 62, 193–217. doi: 10.1037/h0047470

Campitelli, G., Gobet, F., Head, K., Buckley, M., and Parker, A. (2007). Brainlocalization of memory chunks in chessplayers. Intl. J. Neurosci. 117, 1641–1659.doi: 10.1080/00207450601041955

Chung, P. H., and Byrne, M. D. (2008). Cue effectiveness in mitigating postcomple-tion errors in a routine procedural task. Intl. J. Hum. Comput. Stud. 66, 217–232.doi: 10.1016/j.ijhcs.2007.09.001

Derryberry, D., and Reed, M. A. (2002). Anxiety-related attentional biases and theirregulation by attentional control. J. Abnorm. Psychol. 111, 225. doi: 10.1037/0021-843X.111.2.225

Ericsson, K. A., and Kintsch, W. (1995). Long-term working memory. Psychol. Rev.102, 211. doi: 10.1037/0033-295X.102.2.211

Ericsson, K. A., and Lehmann, A. C. (1996). Expert and exceptional performance:evidence of maximal adaptation to task constraints. Annu. Rev. Psychol. 47, 273–305. doi: 10.1146/annurev.psych.47.1.273

Ericsson, K. A., and Towne, T. J. (2010). Expertise. Wiley Interdiscipl. Rev. 1, 404–416.Fajkowska, M., and Derryberry, D. (2010). Psychometric properties of attentional

control scale: the preliminary study on a Polish sample. Polish Psychol. Bull. 41,1–7. doi: 10.2478/s10059-010-0001-7

Farrington-Darby, T., and Wilson, J. R. (2006). The nature of expertise: a review.Appl. Ergon. 37, 17–32. doi: 10.1016/j.apergo.2005.09.001

Finkbeiner, M., and Forster, K. I. (2007). Attention, intention, and domain-specificprocessing. Trends Cogn. Sci. 12, 59–64. doi: 10.1016/j.tics.2007.11.003

Gigerenzer, G., and Gaissmaier, W. (2011). Heuristic decision making. Annu. Rev.Psychol. 62, 451–482. doi: 10.1146/annurev-psych-120709-145346

Goodenough, D. R. (1976). The role of individual differences in field dependenceas a factor in learning and memory. Psychol. Bull. 83, 675. doi: 10.1037/0033-2909.83.4.675

Holyoak, K. J., and Cheng, P. W. (2011). Causal learning and inference as arational process: the new synthesis. Annu. Rev. Psychol. 62, 135–163. doi:10.1146/annurev.psych.121208.131634

Houmourtzoglou, E., Kourtessis, T., Michalopoulou, M., and Derri, V. (1998).Differences in several perceptual abilities between experts and novices in bas-ketball, volleyball and water-polo. Percept. Mot. Skills 86, 899–912. doi:10.2466/pms.1998.86.3.899

Jagiellowicz, J., Xu, X., Aron, A., Aron, E., Cao, G., Feng, T., et al. (2010). The traitof sensory processing sensitivity and neural responses to changes in visual scenes.Soc. Cogn. Affect. Neurosci. 6, 38–47. doi: 10.1093/scan/nsq001

Klein, G. (2011). Critical thoughts about critical thinking. Theoret. Issues Ergon. Sci.12, 210–224. doi: 10.1080/1464536X.2011.564485

Klein, G., Calderwood, R., and Clinton-Cirocco, A. (2010). Rapid decision makingon the fire ground: the original study plus a postscript. J. Cogn. Eng. Decis. Mak.4, 186–209. doi: 10.1518/155534310X12844000801203

Lagnado, D. A., Newell, B. R., Kahan, S., and Shanks, D. R. (2006). Insightand strategy in multiple-cue learning. J. Exp. Psychol. 135, 162–183. doi:10.1037/0096-3445.135.2.162

Logan, G. D. (2002). An instance theory of attention and memory. Psychol. Rev. 109,376. doi: 10.1037/0033-295X.109.2.376





Loveday, T., and Wiggins, M. W. (2014). Cue utilization and broad indica-tors of workplace expertise. J. Cogn. Eng. Decis. Mak. 8, 98–113. doi:10.1177/1555343413497019

Loveday, T., Wiggins, M., Festa, M., Schell, D., and Twigg, D. (2013a). “Pattern recog-nition as an indicator of diagnostic expertise,” in Pattern Recognition – Applicationsand Methods, Vol. 204, eds P. Latorres Carmona, J. S. Sanchez, and A. L. N. Fred(Berlin: Springer), 1–11. doi: 10.1007/978-3-642-36530-0_1

Loveday, T., Wiggins, M. W. Harris, J., Smith, N., and O’Hare, D. (2013b). Anobjective approach to identifying diagnostic expertise amongst power systemcontrollers. Hum. Factors 55, 90–107. doi: 10.1177/0018720812450911

Loveday, T., Wiggins, M. W., Searle, B. J., Festa, M., and Schell, D. (2013c). Thecapability of static and dynamic features to distinguish competent from gen-uinely expert practitioners in pediatric diagnosis. Hum. Factors 55, 125–137. doi:10.1177/0018720812448475

Markovits, H. (2013). Physical aggression facilitates social information pro-cessing. J. Exp. Soc. Psychol., 49, 1023–1026. doi: 10.1016/j.jesp.2013.07.005

Momm, T., Blickle, G., and Liu, Y. (2013). Political skill and emotional cuelearning via voices: a training study. J. Appl. Soc. Psychol. 43, 2307–2317. doi:10.1111/jasp.12180

Morrison, B., Wiggins, M. W., Bond, N., and Tyler, M. (2013). measuringcue strength as a means of identifying an inventory of expert offender pro-filing cues. J. Cogn. Eng. Decis. Mak. 7, 211–226. doi: 10.1177/1555343412459192

Müller, S., Abernethy, B., and Farrow, D. (2006). How do world-class cricket bats-men anticipate a bowler’s intention? Q. J. Exp. Psychol. 59, 2162–2186. doi:10.1080/02643290600576595

Pachur, T., and Marinello, G. (2013). Expert intuitions: how to model the deci-sion strategies of airport customs officers? Acta Psychol. 144, 97–103. doi:10.1016/j.actpsy.2013.05.003

Salthouse, T. A. (1991). “Expertise as the circumvention of human informationprocessing,” in Toward a General Theory of Expertise: Prospects and Limits, eds K.A. Ericsson and J. Smith (Cambridge, NY: Cambridge University Press), 286–300.

Sherbino, J., Dore, K. L., Wood, T. J., Young, M. E., Gaissmaier, W., Kreuger, S., et al.(2012). The relationship between response time and diagnostic accuracy. Acad.Med. 87, 785–791. doi: 10.1097/ACM.0b013e318253acbd

Small, A. J., Wiggins, M. W., and Loveday, T. (2014). Cue – based processingcapacity, cognitive load and the completion of simulated short – duration vig-ilance tasks in power transmission control. Appl. Cogn. Psychol. doi: 10.1002/acp.3016

Weiss, D. J., and Shanteau, J. (2003). Empirical assessment of expertise. Hum. Factors45, 104–116. doi: 10.1518/hfes.45.1.104.27233

Wiggins, M. W. (2006). “Cue-based processing and human performance,” in Ency-clopedia of Ergonomics and Human Factors, 2nd Edn, ed. W. Karwowski, (London:Taylor and Francis), 641–645.

Wiggins, M. W. (2012). The role of cue utilisation and adaptive interface designin the management of skilled performance in operations control. Theor. IssuesErgon. Sci. 1–12. doi: 10.1080/1463922X.2012.724725

Wiggins, M. W. (2014). Differences in situation assessments and prospective diag-noses of simulated weather radar returns amongst experienced pilots. Intl. J.Indus. Ergon. 44, 1, 18–23. doi: 10.1016/j.ergon.2013.08.006

Wiggins, M. W., Azar, D., Hawken, J., Loveday, T., and Newman, D. (2014). Cue-utilisation typologies and pilots’pre-flight and in-flight weather decision-making.Safe. Sci. 65, 118–124. doi: 10.1016/j.ssci.2014.01.006

Wiggins, M. W., Harris, J., Loveday, T., and O’Hare, D. (2010). EXPERT IntensiveSkills Evaluation (EXPERTise) Test. Sydney: Macquarie University.

Wiggins, M., and O’Hare, D. (2003). Weatherwise: evaluation of a cue-based trainingapproach for the recognition of deteriorating weather conditions during flight.Hum. Factors 45, 337–345. doi: 10.1518/hfes.45.2.337.27246

Witteman, C. L., Weiss, D. J., and Metzmacher, M. (2012). Assessing diagnos-tic expertise of counselors using the Cochran–Weiss–Shanteau (CWS) index. J.Counsel. Dev. 90, 30–34. doi: 10.1111/j.1556-6676.2012.00005.x

Wulf, G., McNevin, N. H., Fuchs, T., Ritter, F., and Toole, T. (2000). Atten-tional focus in complex skill learning. Res. Q. Exerc. Sport 71, 229–239. doi:10.1080/02701367.2000.10608903

Young, B. W., and Salmela, J. H. (2010). Examination of practice activities related tothe acquisition of elite performance in Canadian middle distance running. Intl.J. Sport Psychol. 41, 73–90.


Received: 01 April 2014; accepted: 15 May 2014; published online: 03 June 2014.Citation: Wiggins MW, Brouwers S, Davies J and Loveday T (2014) Trait-based cueutilization and initial skill acquisition: implications for models of the progression toexpertise. Front. Psychol. 5:541. doi: 10.3389/fpsyg.2014.00541This article was submitted to Cognition, a section of the journal Frontiers in Psychology.Copyright © 2014 Wiggins, Brouwers, Davies and Loveday. This is an open-accessarticle distributed under the terms of the Creative Commons Attribution License(CC BY). The use, distribution or reproduction in other forums is permitted, pro-vided the original author(s) or licensor are credited and that the original publication inthis journal is cited, in accordance with accepted academic practice. No use, distributionor reproduction is permitted which does not comply with these terms.







OPINION ARTICLEpublished: 16 July 2014

doi: 10.3389/fpsyg.2014.00737

Understanding expertise and non-analytic cognition infingerprint discriminations made by humansMatthew B. Thompson*, Jason M. Tangen and Rachel A. Searston

School of Psychology, The University of Queensland, St Lucia, QLD, Australia*Correspondence: [email protected]


Guillermo Campitelli, Edith Cowan University, Australia

Keywords: expertise, fingerprints, decision making, forensics, non-analytical reasoning, testimony, instance-based learning

As a novice in a particular domain, thecognitive feats that experts are capable ofperforming seem impressive, even extraor-dinary. According to the well-establishedexemplar theory of categorization (e.g.,Brooks, 1987; Medin and Ross, 1989), anew category member in everyday clas-sification (e.g., a bird, a table, or a car)or expert classification (e.g., an abnor-mal chest x-ray, a patient with myocardialischaemia, or a poor chess move) is cat-egorized on the basis of its similarity toindividual prior cases. Often this sensitiv-ity develops effortlessly and without anyintention to learn similarities or differ-ences among the exemplars.

Experts can do a lot with a little. Acrossvarious domains of expertise, it seems thatexperts can perform quickly and accu-rately when given only a small amountof information, as in chess (Gobet andCharness, 2006); fireground command(Klein, 1998); radiology (Myles-Worsleyet al., 1988; Evans et al., 2013), and der-matology (Norman et al., 1989). The expe-riential knowledge based on the hundredsof thousands of prior instances serves as arich source of analogies to permit efficientproblem solving.

A fruitful approach to understand-ing these cognitive feats has been tounderstand where expertise lies in var-ious domains. Expertise in ball sports,for example, seems to lie in anticipatingwhere the ball will be (Abernethy, 1991);expertise in wine seems to lie in apply-ing verbal labels (Hughson and Boakes,2001); expertise in radiology seems to liein rapid discrimination of normal andabnormal radiographs (Evans et al., 2013);and expertise in chess seems lie in rapidretrieval of board configurations frommemory (Chase and Simon, 1973).

Over the last several years, we havebeen working with a fascinating groupof experts who spend several hours aday examining a highly structured set ofimpressions. When a fingerprint is foundat a crime scene it is a human exam-iner, not a machine, who is faced withthe task of identifying the person wholeft it. Professional fingerprint examin-ers are usually sworn police officers whouse image enhancement tools, such asPhotoshop or a physical magnifying glass,and database tools to provide a list ofpossible matching candidates. They placea crime scene print and a suspect printside-by-side—physically or on a computerscreen—and visually compare the prints tojudge whether the prints came from thesame person or two different people.

These fingerprint examiners have tes-tified in court for over one hundredyears, but there have been few experi-ments directly investigating the extent towhich experts can correctly match fin-gerprints to one another, how competentand proficient fingerprint experts are, howexaminers make their decisions, or the fac-tors that affect performance (Loftus andCole, 2004; Saks and Koehler, 2005; Vokeyet al., 2009; Spinney, 2010b; Thompsonet al., 2013a). Indeed, many examinershave even claimed that fingerprint iden-tification is infallible (Federal Bureau ofInvestigation, 1984). Academics, judges,scientists, and US Senators have reportedon the absence of solid scientific prac-tices in the forensic sciences. They high-light the absence of experiments on humanexpertise in forensic pattern matching,suggesting that faulty analyses may becontributing to wrongful convictions ofinnocent people (Edwards, 2009; NationalResearch Council, 2009; Campbell, 2011;

Carle, 2011; Expert Working Group onHuman Factors in Latent Print Analysis,2012; Maxmen, 2012), and they lamentthe lack of a research culture in the foren-sic sciences (Mnookin et al., 2011). Thefield of forensics is, however, beginning toacknowledge the central role that falliblehumans play in the identification process(Tangen, 2013).

Our first point of inquiry was to seewhether qualified, court practicing finger-print examiners are any more accuratethan the person on the street, and to get afeel for the kinds of errors examiners make.In our first experiment (Tangen et al.,2011), we tested the matching accuracyof fingerprint examiners from Australianstate and federal law enforcement agencies.In a signal detection paradigm, we cre-ated ground-truth matching prints for useas targets, and highly-similar, nonmatch-ing prints from a national database searchfor use as distractors. We found that qual-ified, court-practicing fingerprint expertswere exceedingly accurate compared withnovices. Experts tended to err on the sideof caution by making more errors of thesort that could allow a guilty person toescape detection than errors of the sortthat could falsely incriminate an inno-cent person. A similar experiment, withparticipants from the US Federal Bureauof Investigation, produced similar results(Ulery et al., 2011), and a follow-up exper-iment found variability in the consistencywithin and between examiners’ decisions(Ulery et al., 2012). An examiner’s exper-tise seems to lie, not in matching printsper se, but in discriminating highly similarbut nonmatching prints (Thompson et al.,2013a).

In a follow-up experiment (Thompsonet al., 2013b), we replicated (Tangen







http://www.frontiersin.org/journal/10.3389/fpsyg.2014.00737/full







Thompson et al. Expertise in fingerprint discrimination

et al., 2011) but with genuine crime scenematching prints from casework (where theground truth is uncertain), and with theaddition of two trainee participant groups.Intermediate trainees—despite their lackof qualification and average 3.5 years’experience—performed about as accu-rately as qualified experts who had anaverage 17.5 year’s experience. It appearsthat people can learn to distinguish match-ing from similar nonmatching prints toroughly the same level of accuracy asexperts after a few years of experienceand training. New trainees—despite their5-week, full-time training course or their6 months experience—were not any bet-ter than novices at discriminating match-ing and similar nonmatching prints, theywere just more conservative. It appearsthat early training and/or experience maynot necessarily result in more accuratejudgments, but may simply result in amore conservative response bias. Again weconcluded that the superior performanceof experts was a result of their ability toidentify the highly similar, but nonmatch-ing fingerprints as such.

What the findings mean for reason-ing about expert performance in the wildis an open question (e.g., Koehler, 2008,2012; Mnookin, 2008; Thompson et al.,2013a), but these findings do contradictthe notion that fingerprint identification isinfallible, and that the fingerprint identifi-cation “methodology” can be disembodiedfrom a judgment about whether two fin-gerprints match or not.

Our second point of inquiry was tounderstand the nature of expertise in fin-gerprint matching—to understand wherea fingerprint examiner’s expertise lies.Through experience and feedback, anexpert can rapidly retrieve, from memory,previous instances and decisions relevantto the current situation, whereas novicesrely more on formal rules and proce-dures (Brooks, 2005; Norman et al., 2007).Fingerprint examiners claim that careful,deliberate analysis is the basis of the workthat they do (Busey and Parada, 2010),but a hallmark of genuine expertise is theability to accurately perform a domain rel-evant task quickly (Kahneman and Klein,2009). Do fingerprint examiners rely onthe same non-analytic cognitive processesas experts in other domains of expertise?

Busey et al. (2011) found that expertsmove their eyes differently from novices,and Busey and Vanderkolk (2005) foundthat experts performed better than novicesat identifying the matching fragments offingerprints in noise after a short delay.They also found that inverted fingerprintsproduced a delayed N170 event-relatedpotential response in experts but not innovices, suggesting that experts processupright fingerprints configurally (Buseyand Parada, 2010).

In a series of experiments, we furtherexamined the nature of fingerprint exper-tise (Thompson, 2014). We added artificialnoise to all the print pairs, and invertedhalf and kept the other half upright,and found that experts could discriminateprints even when the prints were highlynoisy. Unexpectedly, fingerprint expertsdid not show the classic inversion effectseen in face recognition. We tested theshort term memory of experts and novicesby separating fingerprint pairs in time bya few seconds, and found that experts werebetter than novices at discriminating print,and that experts were far better at discrim-inating similar, nonmatching prints. Wetested the long term memory of expertsand novices by asking them to learn a setof fingerprints to be recognized a few min-utes later, but found no difference in longterm memory accuracy between expertsand novices, and both groups performedaround the level of chance. We tested theability of experts and novices to discrimi-nate prints by presenting them briefly onscreen, and found that experts could accu-rately discriminate prints when presentedfor just 2000 ms, and the largest differ-ence between experts and novices was onsimilar, nonmatching prints. We then fur-ther reduced the stimuli presentation time,and found that experts were more accuratethan novices overall, and that experts hada much better idea about whether a pairof prints match or not in a rapid period oftime. With such short presentation times(i.e., from 250 to 2000 ms), there is littletime to engage in careful, deliberate anal-ysis of the minutiae in a fingerprint imagein order to make accurate decisions.

These findings suggest that fingerprintexperts are capable of making accuratedecisions when the amount of visualinformation in the prints is dramatically

decreased. It is clear that, through expe-rience, experts can learn the regularitiesof matching and nonmatching prints, andrapidly compare new prints to memoryin order to make accurate judgments. Thefindings above are in stark contrast to thecommon and consistent claims in formaltraining, textbooks, and courtroom testi-mony: that fingerprint identification is a“scientific process” that requires careful,thorough analysis in order for judgmentsto be accurate.

Fingerprint expertise is particularlyinteresting because of the sheer amountof experience that examiners have withthe stimuli. Their full-time job, often indepartments that run 24 hours a day, isto visually compare crime scene prints tosuspect and database candidates. It’s dif-ficult to imagine any other domain inwhich there is so much attention andexposure to a highly constrained stimu-lus, where the task is to definitively reportthat two images come from the same, sin-gle source or not. And the stakes are high:innocent people could be wrongly con-victed, and guilty people could be wronglyacquitted.

This vast experience allows experts toresolve information in a print: to cor-rectly regard ambiguous information thatis more consistent with within-source vari-ability as a “match,” and correctly regardambiguous information that is more con-sistent with between-source variability asa “non-match.” An ambiguous mark on afingerprint, for example, can be regardedas signal (i.e., as evidence of a “match”), orit can be disregarded as noise (i.e., as evi-dence of a “non-match”). This kind of pro-cess is undoubtedly operating in novicestoo, but the ambiguity cannot be suffi-ciently resolved unless the examiner hasaccumulated enough matching and non-matching exemplars in memory to pointto one direction or the other. One clearresult of this vast experience is the experts’capacity to disregard, to “see through” theambiguity and surface structure of similarprints and discriminate them accurately.We think that further study of the natureof fingerprint expertise will inform generaltheories for the development of expertise,while also providing an empirical basis forclaims made by expert witnesses in thecourtroom.





Thompson et al. Expertise in fingerprint discrimination

REFERENCESAbernethy, B. (1991). Visual search strategies and

decision-making in sport. Int. J. Sport Psychol. 22,189–210.

Brooks, L. R. (1987). “Decentralized control of catego-rization: the role of prior processing episodes,” inConcepts and Conceptual Development: Ecologicaland Intellectual Factors in Categorization, edU. Neisser (Cambridge: Cambridge UniversityPress), 141–174.

Brooks, L. R. (2005). The blossoms and theweeds. Can. J. Exp. Psychol. 59, 62–74. doi:10.1037/h0087462

Busey, T. A., and Parada, F. J. (2010). The nature ofexpertise in fingerprint examiners. Psychon. Bull.Rev. 17, 155–160. doi: 10.3758/PBR.17.2.155

Busey, T. A., and Vanderkolk, J. R. (2005). Behavioraland electrophysiological evidence for configuralprocessing in fingerprint experts. Vision Res. 45,431–448. doi: 10.1016/j.visres.2004.08.021

Busey, T. A., Yu, C., Wyatte, D., Vanderkolk, J.,Parada, F., and Akavipat, R. (2011). Consistencyand variability among latent print examiners asrevealed by eye tracking methodologies. J. ForensicIdentificat. 61, 60–91. Available online at: https://www.ncjrs.gov/App/Publications/abstract.aspx?ID=255267

Campbell, A. (2011). The Fingerprint Inquiry Report.Edinburgh: APS Group Scotland.

Carle, D. (2011). Leahy Proposes LandmarkForensics Reform Legislation. Availableonline at: http://www.leahy.senate.gov/press/leahy-proposes-landmark-forensics-reform-legislation/

Chase, W. G., and Simon, H. A. (1973). “The mind’seye in chess,” in Visual Information Processing, edW. G. Chase (New York, NY: Academic Press),215–281.

Edwards, H. T. (2009). Statement of the HonorableHarry T. Edwards: Strengthening Forensic Sciencein the United States: A Path Forward. Vol. 47.Washington, DC: United States Senate Committeeon the Judiciary, 1–12.

Evans, K. K., Georgian-Smith, D., Tambouret, R.,Birdwell, R. L., and Wolfe, J. M. (2013). The gistof the abnormal: above-chance medical decisionmaking in the blink of an eye. Psychon. Bull. Rev.20, 1170–1175. doi: 10.3758/s13423-013-0459-3

Expert Working Group on Human Factors in LatentPrint Analysis. (2012). Latent Print Examinationand Human Factors: Improving the PracticeThrough a Systems Approach. Washington, DC:U.S. Government Printing Office.

Federal Bureau of Investigation. (1984). The Science ofFingerprints: Classification and Uses. Washington,DC: U.S. Government Printing Office.

Gobet, F., and Charness, N. (2006). “Expertisein chess”. in Cambridge Handbook of Expertiseand Expert Performance, eds K. A. Ericsson, N.

Charness, P. Feltovich, and R. Hoffman (New York,NY: Cambridge University Press), 523–538.

Hughson, A. L., and Boakes, R. A. (2001).Perceptual and cognitive aspects of wineexpertise. Aust. J. Psychol. 53, 103–108. doi:10.1080/00049530108255130

Kahneman, D., and Klein, G. (2009). Conditionsfor intuitive expertise: a failure to disagree. Am.Psychol. 64, 515–526. doi: 10.1037/a0016755

Klein, G. A. (1998). Sources of Power. Cambridge, MA:The MIT Press.

Koehler, J. (2008). Fingerprint error rates and profi-ciency tests: What they are and why they matter.Hast. Law J. 59, 1077–1100. Available online at:http://ssrn.com/abstract=1486291

Koehler, J. (2012). Proficiency tests to estimate errorrates in the forensic sciences. Law Probabil. Risk 12,89–98. doi: 10.1093/lpr/mgs013

Loftus, E. F., and Cole, S. A. (2004). Contaminatedevidence. Science 304, 959. doi: 10.1126/sci-ence.304.5673.959b

Maxmen, A. (2012). Proposed Bill Calls forBetter Forensic Science. Available online at:http://blogs.nature.com/news/2012/07/proposed-bill-calls-for-better-forensic-science.html

Medin, D., and Ross, B. (1989). “The specific char-acter of abstract thought: categorization, prob-lem solving, and induction,” in Advances inthe Psychology of Human Intelligence, Vol. 5, edR. Sternberg (San Diego, CA: Academic Press),189–223.

Mnookin, J. L. (2008). The validity of latent finger-print identification: confessions of a fingerprint-ing moderate. Law Probabil. Risk 7, 127–141. doi:10.1093/lpr/mgm022

Mnookin, J. L., Cole, S. A., Dror, I. E., Fisher, B. A.J., Houk, M. M., Inman, K., et al., (2011). Theneed for a research culture in the forensic sciences.UCLA Law Rev. 58, 725. doi: 10.2139/ssrn.1755722

Myles-Worsley, M., Johnston, W. A., and Simons,M. A. (1988). The influence of expertise on X-ray image processing. J. Exp. Psychol. Learn. Mem.Cogn. 14, 553.

National Research Council. (2009). StrengtheningForensic Science in the United States: A PathForward. Washington, DC: The NationalAcademies Press.

Norman, G. R., Rosenthal, D., Brooks, L. R., Allen,S. W., and Muzzin, L. J. (1989). The developmentof expertise in dermatology. Arch. Dermatol. 125,1063–1068.

Norman, G. R., Young, M., and Brooks, L. R. (2007).Non-analytical models of clinical reasoning: therole of experience. Med. Educ. 41, 1140–1145. doi:10.1111/j.1365-2923.2007.02914.x

Saks, M. J., and Koehler, J. J. (2005). The comingparadigm shift in forensic identification sci-ence. Science 309, 892–895. doi: 10.1126/sci-ence.1111565

Spinney, L. (2010b). Science in court: the fine print.Nature 464, 344–346. doi: 10.1038/464344a

Tangen, J. M. (2013). Identification personi-fied. Aust. J. Forensic Sci. 45, 315–322. doi:10.1080/00450618.2013.782339

Tangen, J. M., Thompson, M. B., and McCarthy, D. J.(2011). Identifying fingerprint expertise. Psychol.Sci. 22, 995–997. doi: 10.1177/0956797611414729

Thompson, M. B. (2014). On Expertise in FingerprintIdentification (Unpublished Doctoral Dissertation).Brisbane: The University of Queensland.

Thompson, M. B., Tangen, J. M., and McCarthy, D.J. (2013a). Expertise in fingerprint identification.J. Forensic Sci. 58, 1519–1530. doi: 10.1111/1556-4029.12203

Thompson, M. B., Tangen, J. M., and McCarthy,D. J. (2013b). Human matching performanceof genuine crime scene latent fingerprints.Law Hum. Behav. 38, 84–93. doi: 10.1037/lhb0000051

Ulery, B. T., Hicklin, R. A., Buscaglia, J., andRoberts, M. A. (2011). Accuracy and reliabil-ity of forensic latent fingerprint decisions. Proc.Natl. Acad. Sci. U.S.A. 108, 7733–7738. doi:10.1073/pnas.1018707108

Ulery, B. T., Hicklin, R. A., Buscaglia, J., and Roberts,M. A. (2012). Repeatability and reproducibil-ity of decisions by latent fingerprint examiners.PLoS ONE 7:e32800. doi: 10.1371/journal.pone.0032800

Vokey, J. R., Tangen, J. M., and Cole, S. A. (2009).On the preliminary psychophysics of fingerprintidentification. Q. J. Exp. Psychol. 62, 1023–1040.doi: 10.1080/17470210802372987

Conflict of Interest Statement: The authors declarethat the research was conducted in the absence of anycommercial or financial relationships that could beconstrued as a potential conflict of interest.

Received: 15 June 2014; accepted: 24 June 2014;published online: 16 July 2014.Citation: Thompson MB, Tangen JM and Searston RA(2014) Understanding expertise and non-analytic cog-nition in fingerprint discriminations made by humans.Front. Psychol. 5:737. doi: 10.3389/fpsyg.2014.00737This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Thompson, Tangen and Searston.This is an open-access article distributed under theterms of the Creative Commons Attribution License(CC BY). The use, distribution or reproduction in otherforums is permitted, provided the original author(s) orlicensor are credited and that the original publicationin this journal is cited, in accordance with acceptedacademic practice. No use, distribution or reproduc-tion is permitted which does not comply with theseterms.


https://www.ncjrs.gov/App/Publications/abstract.aspx?ID=255267



http://www.leahy.senate.gov/press/leahy-proposes-landmark-forensics-reform-legislation/



http://ssrn.com/abstract=1486291

http://blogs.nature.com/news/2012/07/proposed-bill-calls-for-better-forensic-science.html

http://blogs.nature.com/news/2012/07/proposed-bill-calls-for-better-forensic-science.html











ORIGINAL RESEARCH ARTICLEpublished: 26 November 2014doi: 10.3389/fpsyg.2014.01333

Expert analogy use in a naturalistic settingDonald R. Kretz1 and Daniel C. Krawczyk1,2*

1 School of Behavioral and Brain Sciences, The University of Texas at Dallas, Richardson, TX, USA2 Department of Psychiatry, University of Texas Southwestern Medical Center, Dallas, TX, USA

Edited by:


Reviewed by:

Lindsey Engle Richland, Universityof Chicago, USAMáximo Trench, Universidad delComahue, Argentina

*Correspondence:

Daniel C. Krawczyk, Department ofPsychiatry, The University of Texasat Dallas, 800 W. Campbell Road,GR41, Richardson, TX 75083, USAe-mail: [email protected]

The use of analogy is an important component of human cognition. The type of analogywe produce and communicate depends heavily on a number of factors, such as thesetting, the level of domain expertise present, and the speaker’s goal or intent. In thisobservational study, we recorded economics experts during scientific discussion andexamined the categorical distance and structural depth of the analogies they produced.We also sought to characterize the purpose of the analogies that were generated. Ourresults supported previous conclusions about the infrequency of superficial similarity insubject-generated analogs, but also showed that distance and depth characteristics weremore evenly balanced than in previous observational studies. This finding was likely dueto the nature of the goals of the participants, as well as the broader nature of theirexpertise. An analysis of analogical purpose indicated that the generation of concretesource examples of more general target concepts was most prevalent. We also notedfrequent instances of analogies intended to form visual images of source concepts. Othercommon purposes for analogies were the addition of colorful speech, inclusion (i.e.,subsumption) of a target into a source concept, or differentiation between source andtarget concepts. We found no association between depth and either of the other two

characteristics, but our findings suggest a relationship between purpose and distance; i.e.,that visual imagery typically entailed an outside-domain source whereas exemplificationwas most frequently accomplished using within-domain analogies. Overall, we observed arich and diverse set of spontaneously produced analogical comparisons. The high degreeof expertise within the observed group along with the richly comparative nature of theeconomics discipline likely contributed to this analogical abundance.

Keywords: analogy, expertise, naturalistic, reasoning, problem-solving

INTRODUCTIONThe importance of analogy has been described in many ways bynotable researchers. Polya (1957) wrote that analogy “pervadesall our thinking,” Holyoak et al. (2001) called it “a central com-ponent of human cognition,” and Hofstadter (2001) referred toit as “the lifeblood . . . of human thinking.” Because of its per-ceived importance in cognitive functioning, the use of analogy inthought and language has been studied extensively in cognitivepsychology, cognitive development, and cognitive science sincethe early 1980s. Research examining analogical production andretrieval under experimental conditions has provided a wealth ofvaluable data. Far fewer studies of these important phenomenahave been conducted in naturalistic settings.

The importance of studying analogical production “in thewild” was emphasized by one of its most prominent researchers.Just as it is necessary to conduct both in vivo and in vitro stud-ies to fully understand biological phenomena, Dunbar (1995,2001) argued that it is likewise necessary to conduct bothnaturalistic and experimental studies in cognitive research tofully understand the cognitive processes involved in reason-ing and analysis. He modeled his approach after techniquesapplied in biological research, referring to this paradigm as

“in vivo/in vitro.” Observing behavior in naturalistic settingsprovides several advantages: (a) behaviors are unconstrained bylaboratory conditions and are not instigated by artificial or exper-imental stimuli, (b) the setting emphasizes processes rather thanoutcomes, and (c) there is a clear relationship between observedbehaviors and the domain of interest (Dunbar, 1995; Cranoand Brewer, 2002). These conditions are particularly importantwhen investigating analogical thinking which involves linkingone’s current context with prior knowledge in a spontaneousfashion. It is important to note, however, that the observationalapproach lacks the superior structure and clarity of laboratory-based experiments.

Because there are many commonly used definitions of the termanalogy, we felt it necessary to offer a clear definition that capturesthe essential characteristics applied by researchers in this field.A frequently used means of conveying an understanding of anunfamiliar concept is by drawing a comparison to similar, morefamiliar concepts. An analogy conveys more than literal similaritybetween two objects or concepts (Gentner, 1983). As a process,analogy involves the search for and selection of a well-understoodsource from long-term memory, followed by the transfer of mean-ing from that source to a less familiar target (Spellman and

www.frontiersin.org November 2014 | Volume 5 | Article 1333 | 108













Kretz and Krawczyk Analogy in a naturalistic setting

Holyoak, 1996). The set of correspondences between a source andtarget are referred to as a source-target mapping. In contrast toother forms of likeness or similarity, analogy is based on a com-parison of structural relations, or systems of relations, rather thana mere resemblance of surface properties or attributes (Gentnerand Markman, 1997). A system of relations can be quite complex,and all of its mappings may not be apparent. Some evidence sug-gests, however, that analogical mapping is not entirely relational(Ball et al., 2004; Bearman et al., 2007).

In the current study, we applied Dunbar’s in vivo approachto observe and analyze some of the characteristics of sourceanalogs produced during live, open discussion involving prob-lems in economics and discuss some of the factors influencingtheir production. We chose a behavioral economics group toobserve, as economists seek to broadly explain complex real-world human behavior in the context of models, games, andexamples. All of these techniques rely heavily upon comparisonsbetween different states of the world, different types of behav-ior, and combinations of models and situations. Many of thesecomparisons are potentially analogical, drawing heavily uponthe expertise of the individual and the broader group. Datawere collected from the weekly sessions of an economics read-ing group, which consisted of participants whose expertise in thebroad field of economics varied in terms of depth and academicspecialization.

ANALOGICAL DEPTHOne of the factors that has been shown to influence sourceretrieval is the level, or depth, of similarity between a target andthe chosen source. Past research has shown that source-to-targetmapping is primarily driven by the comparison of structuralrelations between source and target concepts but retrieval maybe facilitated by superficial characteristics (Gick and Holyoak,1980, 1983; Gentner et al., 1993; Forbus et al., 1997). In otherwords, people will rely on structural details to find appropriatesources when mapping an analogy, but perhaps find it easier torely on overlapping surface features in addition to underlyingstructure when retrieving a correspondence (Holyoak and Koh,1987; Holyoak and Thagard, 1995; Dunbar, 2001). The use ofsuperficial characteristics for comparison is particularly apparentwhen source and target analogs are generated a priori and pro-vided to an individual, who must then consider the nature of therelationship or relationships.

When people generate their own analogies by drawing upontheir own knowledge, research has shown that fewer super-ficial similarities between analogs are observed. Studies byDunbar (1995, 1997) and Blanchette and Dunbar (1997, 2000,2001) revealed that more than half of the analogies pro-duced during biology laboratory meetings and political dis-course, respectively, showed no apparent surface similarity.The Bearman et al. (2007) management study showed that73% of analogies were only structural in nature. With thisin mind, the first goal of this study was to examine thedepth of analogies produced by economics experts during sci-entific discourse. Results were expected to show infrequent sur-face feature overlap and provide additional support for earlierfindings.

ANALOGICAL DISTANCEA second important feature in analogy use concerns the range, ordistance, between the domains of the source and target analogs.Dunbar (1995) defined three categories of analogy in terms oftheir degree of domain separation: (a) long-distance describes asource drawn from a very different domain (also referred to asoutside-domain, or out-of-category), (b) regional refers to a sourcemapped from a similar domain (e.g., economics to finance orpublic administration), and (c) local maps a target to a sourcein the same domain. Both local and regional classes are collec-tively referred to as within-domain, or within-category. Because ofthe subjective nature of judging domain distance between similardomains, many observational studies simply categorize analo-gies using a binary choice of within-category or outside-category.Analogies formed within the same category might compare bio-logical organisms or investment strategies, while commonly refer-enced domains for outside-category analogies include sports (e.g.,one might fail and “strike out at the plate” or succeed and “pushthe ball over the goal line”) and the supernatural (e.g., “it worksso well that it’s like magic”).

Research on the domain separation of analogies has pro-vided contrasting results. Dunbar’s studies of scientific reasoningset in microbiology laboratories showed heavy use of within-category analogies—98% of the analogies generated were clas-sified as either local or regional (Dunbar, 1995, 1997). A studyby Saner and Schunn (1999) produced a similar but narrowerfinding—researchers in psychology laboratory meetings and col-loquia made frequent use of analogies, and more than 75% ofthem were within the same domain. In contrast, Blanchette andDunbar found that 77% of the analogies that appeared in politicalarticles aimed at more general audiences were outside-category(2001). Christensen and Schunn’s, 2007 engineering design groupstudy showed a more balanced mix of analogies—55% werewithin-category while 45% were outside-category.

One explanation for the difference in domain distance is thatthe collective expertise of the audience may influence the selectionof analogy. When addressing other domain experts with special-ized knowledge, within-category analogs may prove more effec-tive whereas an outside-domain analogy might be more attractivechoice for a general audience. The type of task or function per-formed by the analogizer is another possible constraint. Oursecond goal was to observe the use of within- and outside-domainanalogies produced by the economics experts. We expected tofind a fairly balanced use of both styles, which differs somewhatfrom earlier findings. Our expectation was motivated by the dif-ferent subdomains and varying levels of subject expertise withinthe reading group. Furthermore, in light of the stated belief thatwithin-domain analogies tend to involve a higher degree of super-ficial similarity than outside-domain analogies, we investigatedwhether such a relationship emerged.

ANALOGICAL PURPOSEThirdly, past research has shown that the goal of the individ-ual producing the analogy is likely to influence the process ofselection and transfer (Spellman and Holyoak, 1996). Prior obser-vational studies have examined the types of goals that emergefrom the production of analogies, and the goals themselves

Frontiers in Psychology | Cognition November 2014 | Volume 5 | Article 1333 | 109





tend to be highly domain-specific and task-specific. For exam-ple, the four goals derived from the microbiology laboratoriesby Dunbar (1995, 1997) were: forming hypotheses, designingexperiments, modifying experiments, and explaining issues andconcepts to other scientists. The management decision-makingstudy by Bearman et al. (2002) identified only two goals: problem-solving and illustration. Similarly, the Christensen and Schunn(2007) study of analogizing during engineering design meetingsreported three functions: identifying problems, solving problems,and explaining concepts. Our economics reading group differedin its function from the settings of the studies mentioned above.The purpose of the group was limited to explaining and under-standing experiments performed by other researchers; i.e., theexperts in our observational study critiqued experimental designsand analyzed results.

Blanchette and Dunbar (2001) provided evidence that goalsinfluenced the choice of analogy. When supporting a favor-able position, emotionally positive analogies were more com-monly chosen over those displaying negative emotional ideas.Conversely, when criticizing an unfavorable position, emotion-ally negative analogies were more frequent. Because analogicalretrieval appears subject to the influence of the purpose for whichit was produced, our third goal was to analyze the range of goalsthat emerge from the use of analogies by the economics group andtheir potential effect on analogical distance and depth.

METHODSPARTICIPANTSThe setting for this study was the School of Economic, Political,and Policy Sciences (EPPS) weekly reading group at TheUniversity of Texas at Dallas during the Spring semester of 2011.Participants included male and female reading group attendees.The average weekly attendance was approximately 20 partici-pants, which included both faculty members (∼25%) and gradu-ate students (∼75%). The group was largely the same from weekto week, but attendance did fluctuate and not all participantsattended regularly. Participants’ academic expertise was mixed,and included sub-disciplines such as econometrics, experimen-tal economics, and game theory. A few of the attendees had noexposure to experimental methods.

The sessions were conducted in typical reading group fash-ion. One of the senior faculty members acted as the group’smoderator, and one student participant was assigned each weekto lead the following week’s discussion of one or more chosenresearch papers. Following a brief presentation of the paper by theassigned student, a free and open discussion followed in which themembers of the group examined and dissected the experimentalmethods and results described in the paper.

PROCEDURESession recordingsThe investigator attended and made audio recordings at each offive group meetings, but did not participate in the discussions andappeared to have minimal impact on group interaction. Groupmembers understood that their discussions would be evaluated(the moderator referred to it as “the science behind the science”).The research goal of examining the spontaneous use of analogies

was not revealed, however. Although each session was scheduledto last for 90 min, actual discussion times ranged from 65 to110 min. In all, approximately 7 h of discourse were recorded.

Transcription procedureFour transcribers participated in the initial processing ofthe discourse—the primary author and three undergraduatePsychology students. Only the primary author and one of theundergraduates had any significant transcription experience priorto the task. None of the undergraduate transcribers had expo-sure to Economics beyond introductory coursework. Transcriberswere given instruction by the primary author on what consti-tutes an analogy based on the definitions presented earlier in thispaper, then solved practice problems on recognizing analogiesfrom non-analogies by identifying sources and targets. The periodof instruction lasted approximately 30 min.

For purposes of indoctrination, each undergraduate tran-scriber was given one of the sessions for practice and directed toprocess at least 30 min of the recording. They were instructed totranscribe passages in which a possible analogy was made, takingcare to include all of the important source and target informa-tion. Their instruction was: “when in doubt, include it.” Theprimary author evaluated their performance and made individualadjustments until the results were satisfactory and consistent.

To address the possibility of subtle, easily-overlooked analo-gies, every audio recording was processed by two transcribers.To be included for analysis, a passage needed only to appear oneither transcriber’s log, not both. In order to be missed, a pas-sage would have to have been heard by both transcribers andrejected.

Coding procedureThe two authors performed the coding duties. Both were expe-rienced coders with strong knowledge of analogy literature andconsiderable research experience. Each of the transcribed seg-ments was first evaluated for the presence of one or more analo-gies based on the definition stated earlier. Comparisons basedonly on literal similarities were considered non-analogies. If a seg-ment was judged to contain no analogy, no further evaluation wasperformed. If an analogy was deemed present, the source, target,source domain, and target domain of the analogy were recorded.Each analogy was then coded along the three dimensions ofdistance, depth, and purpose.

Both coders rated the entire set and reliability was calculatedfor each dimension. Distance was rated using the commonlyapplied within-domain and outside-domain categories. Followingthe example of Saner and Schunn (1999) in which the authorscollapsed all psychology-related categories into “within-domain,”we likewise considered all analogies related to economics, finance,statistics, probabilities, and game theory to be in the samedomain. Depth was rated as superficial if the source and targetshared surface characteristics; if not, it was rated as structural.Two passes were made by each coder to rate the purpose. Thefirst pass was used to generate and agree on a list of functionsrepresented by the set of analogies. Then, the coders used the listfrom the first pass as set of categories for rating analogies in thesecond pass.





RESULTSSEGMENT TRANSCRIPTION AND ANALOGY EXTRACTIONTranscription of the five recorded sessions yielded 114 uniquepassages with possible analogies (M = 16.29 analogy segmentsper hour). See Appendix A for a sampling of extracted pas-sages. When judging whether a passage contained an analogy,the coders agreed on 96% (n = 109) of the 114 passages. Ofthese, most were rated as containing an analogy (n = 91, 83%).Passages lacking the basic ingredients of an analogy (i.e., source-target mapping and knowledge transfer) or showing evidence ofa literal comparison (n = 18, 17%) were eliminated. Five pas-sages were inconclusive and were likewise eliminated. The inter-rater reliability for identifying analogies was strong (κ = 0.85).Some of the passages were found to contain multiple analo-gies. In all, 97 analogies were extracted for the remainder of thisanalysis.

ANALOGICAL DEPTHIn coding for analogical depth, the coders agreed on 91%(n = 88) of the 97 analogies. Of these, analogies showing noobvious overlap in surface characteristics (n = 69, 78%) faroutnumbered those where superficial similarities were present(n = 19, 22%). The inter-rater reliability was found to besubstantial (κ = 0.75).

ANALOGICAL DISTANCEIn coding for analogical distance, the coders agreed on 86%(n = 83) of the 97 analogies. Of these, within-domain (n = 44,53%) and outside-domain (n = 39, 47%) analogies were almostevenly distributed. The inter-rater reliability was again found tobe substantial (κ = 0.71).

ANALOGICAL PURPOSEAt stated above, we did not impose a priori categorical restric-tions on coding for analogical purpose. Rather, the coders werefree to make subjective judgments as to the intent of the speakeron the first pass through the data. From these impressions, wegrouped similar items and derived a set of categories for codingthe analogies on the second pass. The derived taxonomy of ratingcategories was as follows: Differentiation (highlighting differencesbetween source and target), Inclusion (indicating that the targetwas a type or component of the source), Example (indicating thatthe target was an instance of the source concept), Visualization(intended to create a picture or image in the mind of the audi-ence), Emotion (appealing to feelings of the audience, or drawingon the emotion of the expert), Critique (using the source to pointout shortcomings in the target), Exaggeration (gratuitous use ofcolorful phrasing), and Abstraction (broaden the target conceptusing a more general source concept). Where an analogy plausiblyserved multiple purposes, the raters chose the strongest.

In coding for analogical purpose, the coders agreed on 79%(n = 77) of the 97 analogies. Examples were the most prevalent(n = 27, 35%), followed by visualizations (n = 19, 25%). A fairnumber of exaggerations (n = 11, 14%), inclusions (n = 9, 12%),and differentiations (n = 8, 10%) were observed, but abstrac-tions (n = 2, 3%) and critiques (n = 1, 1%) were uncommon.

The inter-rater reliability for analogical purpose was substantial(κ = 0.74).

ASSOCIATION BETWEEN DEPTH AND DISTANCEAs reported above, most of the analogies were rated as struc-tural analogies. Of those, there was an even distribution ofoutside-domain (n = 31, 50%) and within-domain (n = 31,50%) analogies. The superficial analogies were likewise splitbetween outside-domain (n = 7, 54%) and within-domain (n =6, 46%). It was determined that there was no significant distance-depth association [χ2

(1,N = 75) = 0.06, p = 0.80], suggesting thatthe two variables are independent.

ASSOCIATION BETWEEN DEPTH AND PURPOSEAlmost half of the analogies were rated as either structural-example (n = 18, 25%) or structural-visualization analogies (n =16, 23%). Though the distribution of purpose ratings was highlyskewed, it was skewed similarly over both categories of analogi-cal depth and no significant depth-purpose association was found[χ2

(7,N = 71) = 8.74, p = 0.27]. This finding suggests that thesetwo variables are also independent.

ASSOCIATION BETWEEN DISTANCE AND PURPOSEIn contrast to the two findings above, an examination of analog-ical distance as a function of the speaker’s purpose did reveal asignificant association [χ2

(7,N = 66) = 34.94, p < 0.005]. When aspeaker sought to produce an analogy that created a visual image,an out-of-domain source was typically selected (n = 16, 24%).On the other hand, when an analogizing by example, a source-target transfer from within the same domain was most commonlyobserved (n = 18, 27%).

DISCUSSIONThe most prominent observation to come from this study wasthat the use of analogy for exploration and explanation amongeconomics experts and students was rich and abundant. In427 min of discourse, 97 analogies were extracted, suggesting thatanalogies are both commonly used and serve as an importantcomponent in human reasoning and in understanding prob-lems. The reading group setting enabled the observation of actualexperts spontaneously forming analogies using their semanticknowledge applied to economics, a domain likely to entail morefreedom to move between and among domains of knowledge thanthe previously investigated biology and political domains. Theselection and fluid assembly of analogies during discourse mayhelp to reveal the core principles involved in analogical think-ing among experts. This study’s findings will be discussed in thecontext of prior evidence.

EXPERTS IN THE FIELD OF ECONOMICSWe chose to study behavioral economics experts in this study.Economists seek to broadly explain complex real-world humanbehavior in the context of equations, models, games, and hypo-thetical examples. All of these techniques rely heavily upon com-parisons between different states of the world, different types ofbehavior, and combinations of models and situations. While somefeature or object-based similarity occurs in comparisons betweeneconomic models and real-world behavior, it can be argued that






the majority of the comparisons are about relations and sys-tems of relations. For example, in the economic Ultimatum Game(Guth et al., 1982), two individuals each make monetary deci-sions that will be reviewed and reacted to by the other. TheUltimatum Game can be compared to many situations in lifein which banks, individuals, or nations either choose to coop-erate or not with regard to money, goods, or military action.Thus, many of the comparisons used in behavioral economics arelikely to be analogical and may occur with greater frequency thanobserved in laboratory settings. The expertise of the individualsinvolved in behavioral economics discussions and expertise levelsof the broader group are likely to encourage potentially rich andrelationally deep analogies.

Expertise was a critical element to this study, as the roleof domain knowledge in analogy production is not yet wellunderstood. The level of expertise among reading group partic-ipants varied, but all had sufficient knowledge to be consideredexperts in the field. In some observational studies, expertise wasnarrowly concentrated [Dunbar’s (1995, 1997) molecular biol-ogy group], whereas in others, general audiences possessed littleto no domain knowledge [e.g., Blanchette and Dunbar (2001)political news study]. In our economics reading group, the exper-tise was both deep and broad, covering a range of specializa-tions such as experimental economics, econometrics, and gametheory.

There is growing evidence that the depth of expertise of theaudience may influence analogical selection. It is reasonable tohypothesize that the breadth of expertise may contribute, as well.Blanchette and Dunbar (2001) showed more frequent use ofoutside-domain analogies when the intended audience consistedof non-experts and correspondingly more prevalent within-domain analogies among fellow experts. Additionally, expertsappear more likely to exploit the complex relational nature ofanalogies while avoiding the tendency to rely on superficial fea-tures for comparison. The findings of Bearman et al. (2007)suggested that experts tend to compare structural relationshipsin all activities but novices show differences; i.e., when engagedin problem solving activities, novices rely on structural analogiesbut incorporate superficial features in their comparisons whenillustrating or explaining.

We based our depth and distance expectations for this studylargely on these findings. We did, as expected, observe a more evendistribution of distance in analogies and infrequent use of super-ficial features in making comparisons. Relative to novices, expertscan draw on a great deal of accumulated knowledge when think-ing and reasoning (Bearman et al., 2007). It stands to reason thatthey are better able to exploit the deep, structural nature of sourceinformation as a result. Non-experts, on the other hand, lack thesame deep encoding of domain information and may need to relymore heavily on superficial characteristics. The evidence is incon-sistent, however; the Blanchette and Dunbar (2000) study pointsto novice use of structure as well.

ANALOGICAL DEPTH—SUPERFICIAL VS. STRUCTURALStructural correspondences in the underlying system of relationsbetween source and target elements represent an important com-ponent of analogies. Past experimental research demonstrated

that superficial features influence the selection of source analogs(Gick and Holyoak, 1980, 1983; Gentner et al., 1993; Forbuset al., 1997), and studies in naturalistic settings suggest that otherfactors may contribute, as well (Blanchette and Dunbar, 2001;Bearman et al., 2007). The ratings from this study showed thatthe economics experts relied primarily on structural componentsof analogies (78%) with infrequent use of superficial featurecomparisons (22%).

The data collected from the reading group sessions contained adense set of complex comparisons rich in structure. Because of theunconstrained, yet guided, nature of the discussion, participantsexperienced a great deal of freedom to explore, compare, andexplain complex target concepts. The reading group setting hadcertain features of prior story-based analogical reminding studies(Gentner et al., 1993; Wharton et al., 1994; Catrambone, 1997),as participants compared the contents of journal papers and thevarious experimental methods they described. The setting alsohad characteristics of the Blanchette and Dunbar “productionparadigm” studies (2001) in which participants generated sourceanalogs from their own knowledge and experiences. Perhaps notall settings are completely retrieval-based or production-based;rather, the degree to which the activity combines retrieval taskswith production tasks may determine the balance of analogicaldepth applied. Additionally, the occasional use of superficial com-parisons likely reflects a tendency to spark heightened interest bymaking a link to a distant analogous domain during their descrip-tions of economic processes. Indeed, some of the surface-levelanalogies tabulated could be considered to be turns of phrase ormetaphorical comparisons. The use of such comparisons can behelpful in making speech more interesting to the listener and toadd points of common reference periodically.

ANALOGICAL DISTANCE—WITHIN DOMAIN VS. OUTSIDE DOMAINOne possible explanation for the observed balance betweenwithin-category and out-of-category analogies was that subjectmatter expertise shared among a speaker and audience influencedsource selection. Observational data offer support to this hypoth-esis. In Dunbar’s (1995, 1997) biology laboratory study, scientistsoverwhelmingly produced within-domain analogies (98%). Sanerand Schunn (1999) also observed high rates of within-domainanalogies in their study of psychology laboratory meetings (81%within-domain) and colloquia (77% within-domain). In con-trast, Blanchette and Dunbar (2001) studied opinion articles inthe mainstream press, where the audience was the population atlarge. Here, they found that most analogies (77%) were from out-side the target domain. Based on these findings, it appears thatexperts produce more within-domain analogies when commu-nicating knowledge to their fellow subject matter experts, butgenerate analogies from sources from other domains when theaudience consists of non-experts.

A second, but related, explanation stated that the goals ofthe participants were an influencing factor on source distance(Dunbar, 1995; Holyoak and Thagard, 1997; Blanchette andDunbar, 2001). In Dunbar’s biology lab study, the scientists wereheavily focused on examining unexpected experimental resultsand resolving methodological problems. However, in the Dunbarand Blanchette studies involving both experts and non-experts,





the experts’ goals were to educate and/or persuade. In our study,there was no “discovery task”—i.e., there were no new hypothesesbeing generated, no new designs being developed, and no prob-lems being solved. Rather, the discussion involved a great deal ofcomparison, both integrating and differentiating details of exper-imental methods and results. It may be that tasks involving agreater amount of creativity or innovation lead to more frequentuse of same-domain analogies.

The results from the current study were more balanced. Out-of-category analogies were observed most frequently (53%), butwithin-category analogies accounted for a sizeable portion of thetotal as well (47%). These results fall squarely between the pre-vious findings, but a plausible explanation can be made. If thenumber of within- and outside-domain analogies is a functionof expertise, then the balance between them might be expectedto vary with the range and depth of expertise. In the EPPSreading group, the participants all had some degree of expertisein economics, but their domain experience varied in terms ofsub-discipline (e.g., econometrics, game theory, behavioral eco-nomics) and academic career longevity (i.e., faculty or graduatestudent). Hence, common expertise is likely to be an influenc-ing factor, as Dunbar suggested, but the amount of influence ithas on source distance may moderated by the variability in suchfactors as range and depth. Additionally, the goals of the partici-pants differed from those in earlier studies. Here, the participantssought to comprehend papers describing experimental methodsand results, often by comparing unfamiliar methods to known,more familiar ones. To accomplish these goals, perhaps a morebalanced and diverse set of analogies is most effective.

Another possibility for explaining the balance between within-and outside-domain analogies that we observed is concerns theeconomics discipline itself. In Dunbar’s (1995, 1997) biologylab observations, the overwhelming number of within-domainanalogies may have stemmed from the fact that a majority of situ-ations in molecular biology are likely to have clear relational cor-respondences to closely related areas within biological research,rather than remote domains comprised of non-molecular ele-ments. Meanwhile, the Blanchette and Dunbar (2001) study ofpolitical commentary suggested that politicians and journalistsappeared to draw intentionally from remote domains that wouldbe familiar to readers in ways intended to highlight certain aspectsof relational comparisons. In the present study, discussions acrossthe relatively broad domain of economic inquiry highlightedthe technical overlap between its various subdomains. Unlikemolecular biology, economics has a high potential for relationalalignment to more remote domains within public policy, banking,and corporate practice, domains in which human behavior, mon-etary valuations, and world affairs converge. Thus, the complexityof economics, which plays out both in academic analytic settingsand in real-world financial markets, appears to provide a richfield optimal for both within-domain and out-of-domain ana-logical comparisons. Given the complicated nature of economicsystems, economic analogies are rarely complete, so both superfi-cial and structural correspondences appear to be drawn upon inorder to explain and describe various aspects of complex systems.The validity or appropriateness of analogies in economics mayalso be subject to greater interpretation than other domains given

our limitations in fully explaining human behavior and marketdynamics.

In terms of a possible relationship between distance anddepth, it has been suggested that within-domain analogiespresent more superficial similarity than distant analogies; inother words, the greater the distance, the less superficial thecomparison (Christensen and Schunn, 2007). However, wefound no evidence of that constraint in our observation ofeconomists.

ANALOGICAL PURPOSEIt is fairly well established that goals influence the productionof analogies (Dunbar, 1995, 1997; Spellman and Holyoak, 1996;Blanchette and Dunbar, 2001). Prior studies examined the goalsof experts in scientific laboratories in both discovery (e.g., prob-lem solving, hypothesis generation, experimental design) andnon-discovery (e.g., explanation, illustration, or visualization)activities. The primary difference in purpose between the eco-nomics reading group and groups observed in the cited studieswas the absence of discovery goals in the reading group. Since wedetermined that the discussions we observed were largely non-discovery in nature, we focused on categorizing the extractedanalogies into groupings based on the perceived reason for select-ing the particular analogy; e.g., to differentiate a concept fromother concepts, to inject emotion or colorful language into a com-parison, to give a concrete example of a more abstract idea, etc.The list we derived can be found in the Results section above.

Blanchette and Dunbar (2001) provided evidence that goalsinfluence analogical production, but the goals of the individualsin their study appeared to have no effect on analogical distance.Saner and Schunn (1999), on the other hand, found that goals didimpact the domain distance. In particular, they found that indi-viduals used within-domain analogies when working to identifyproblems but used outside-domain analogies when explainingissues or concepts to their lab mates. In our study, too, thepurpose-distance effect was significant. Exemplification was asso-ciated with within-domain analogizing, while visualization wasstrongly related to out-of-domain analogy use. It may be that thefunctions that result in within-domain analogies (i.e., problemidentification, generation of concrete examples) share some ofthe same underlying cognitive operations, as do those that pro-duce outside-domain analogies (i.e., explaining issues, visualizingconcepts), in ways that influence the generation and retrieval ofanalogies.

In interpreting our results, we should mention that the cate-gories we developed were not mutually exclusive; e.g., an analogyintended to differentiate between concepts might do so usingvisual elements. In cases where an analogy plausibly served multi-ple purposes, the raters chose the category they felt best describedits purpose.

GENERAL DISCUSSION ON THE RICHNESS OF OBSERVED ANALOGIESWe have already emphasized that the collected passages con-tained a wealth of analogies rich in depth and structure, many ofwhich involved implied systems of complex relations that couldnot be fully identified and analyzed. Furthermore, some of themore complex comparisons actually involved multiple analogies






at different distances and depths. The challenges we faced interms of analyzing the passages were made more difficult due tothe complexities of the structural and superficial analogies used.Unlike laboratory tasks, in which comparisons between explicitstatements are made (Gentner and Landers, 1985; Gentner et al.,1993; Krawczyk et al., 2004, 2005), the comparisons made by oureconomics experts were not always made fully explicit. Indeed,we frequently encountered instances in which a source or a targetwere implied, or not even articulated in the flow of discussion.This interpretive challenge may have been exacerbated by thehigh degree of expertise in our sample group, as some analo-gies involved inside knowledge that the speaker did not feel wasnecessary to clearly articulate in order to meaningfully draw thecomparison to the group.

These observations are offered for several reasons. First, thereare additional questions that can be investigated from these data,aside from the boundaries of this paper (e.g., Do the within-category/outside-category comparisons correspond to particulartypes of source-target pairs? Does a taxonomy of source categoriesemerge from the outside-category analogies in this setting? Howeffective did the analogies seem to be in conveying the intendedinformation?). Second, the rating process involved some inherentsubjectivity; even though inter-rater reliability was rather high,there were disagreements on specific passages. This is a chal-lenge inherent in real-world analogical analyses, as the dynamicand unscripted nature of the interactions can produce text thatis very difficult to interpret outside of the context in which itwas spoken and by an individual who does not have access tothe speaker’s intent. Third, the rating process was further com-plicated when analogies were embedded in familiar languageconstructs—well-known clichés, common metaphors, etc.—andwere not immediately identified as analogies by the raters, as suchphrases have become so conventionalized that their figurativequalities can be simply overlooked.

We were able to ascertain some of the major purposesfor which analogies are used in economics discussions amongexperts. The leading purpose for analogical comparisons was toprovide examples. Good examples from other domains or famil-iar sources can provide clarity to a target and make the conceptmore concrete. In many of the analogies, we observed that expertstended to either relate an abstracted theoretical model to a morecommon situation that occurs in human interactions, or theydid the reverse and described a particular situation as being anexample of what is known to occur within a particular theoreti-cal game. Experts also made common use of analogies to createa visual image of a particular concept. Visualization is importantfor providing a common ground for the audience and for makinga target domain richer and easier to comprehend. Experts alsoused analogies in order to add color or interest to their contri-butions and to mark a topic as being included within a morefamiliar source domain. Given that we conducted this observationprocess in an academic setting, our experts may have developedtendencies to use analogy due to their experiences with teaching,in which good examples and visual comparisons can be use-ful for conveying concepts (Richland et al., 2007; Glynn, 2008).These goals for analogical comparison fit with prior observationsabout analogy as being diverse in function (Holyoak and Thagard,

1995), but contrary to several prior scholarly works describinganalogical purpose (Holyoak and Thagard, 1997; Hummel andHolyoak, 1997) we did not find substantial evidence of new infer-ences being generated. Rather, the most prevalent use of analogyin the economics experts was to describe concepts to the group orto point out similarities between relational systems in either thereal world, or in theoretical constructs.

In addition to these questions, several limitations with regardto the method and analysis of this study should be noted. Thetranscribers were non-experts in economics, but they were cho-sen to avoid domain bias and reduce the tendency to add theirown interpretation to the passages. Additionally, three of the fourtranscribers lacked prior transcription experience in recognizinganalogies, which opened the door for subtle analogies to be over-looked. We addressed this limitation using a two-phased trainingprocess and a two-person extraction strategy and believe that wereduced - but may not have eliminated - the likelihood of missedanalogies. While it is possible that some subtle analogies may havebeen overlooked, we believe the number to be small enough to notalter our findings.

Their lack of domain expertise, however, may have con-tributed to the subjectivity and inconsistency mentioned above.Furthermore, we did not code the analogies according to anysort of standard domain taxonomy, as Dunbar (1995) did in hismicrobiology laboratory study. Finally, we did not account forindividual differences by examining patterns of analogy use byspecific individuals in the group.

CONCLUSIONSIn summary, this study reinforces the strong reliance humansplace on analogies for developing understanding and communi-cating in natural settings. It contributes valuable evidence thathumans are quite agile in their selection of analogies, drawingon a mix of shallow and deep comparisons and determining aneffective distance strategy based on the constraints of the domainand the level of perceived group expertise. Economics faculty andgraduate student experts engaged in scientific discussion wereobserved to apply analogies that were more balanced in termsof categorical distance and structural depth than those observedin other natural settings. The domain context, problem-solvinggoals, and participant expertise of this particular group settingall appear to be important factors that led to differences in themagnitudes of depth and distance observed in the earlier stud-ies. No evidence of an association between the distance and depthcharacteristics was found, but distance and purpose appeared tobe related. A number of other questions remain to be answeredabout the ways in which analogies are applied in different types ofsettings. Future naturalistic analogy research could help to clar-ify the role on analogical comparisons in real-world settings andthe social sciences appear to provide particularly rich domains foradditional studies.

ACKNOWLEDGMENTSThe authors wish to thank Dr. Catherine Eckel and the Economic,Political, and Policy Sciences (EPPS) Spring 2011 weekly readinggroup at The University of Texas at Dallas for their participationin this study. Thanks, also, to Brandon Wolfe, Srikant Chari, and





Ashley Fournier for their valuable assistance with the difficult taskof transcription.


REFERENCESBall, L. J., Ormerod, T. C., and Morley, N. J. (2004). Spontaneous analogising in

engineering design: a comparative analysis of experts and novices. Des. Stud. 25,495–508. doi: 10.1016/j.destud.2004.05.004

Bearman, C. R., Ball, L. J., and Ormerod, T. C. (2002). “An exploration of real-world analogical problem solving in novices,” in Proceedings of the 24th AnnualConference of the Cognitive Science Society (Mahweh, NJ: Erlbaum).

Bearman, C. R., Ball, L. J., and Ormerod, T. C. (2007). The structure and function ofspontaneous analogising in domain-based problem solving. Thinking Reasoning13, 273–294. doi: 10.1080/13546780600989686

Blanchette, I., and Dunbar, K. (1997). “Constraints underlying analogy use in areal world context: politics,” in Proceedings of the 19th Annual Conference of theCognitive Science Society (Mahweh, NJ: Erlbaum).

Blanchette, I., and Dunbar, K. (2000). How analogies are generated: theroles of structural and superficial similarity. Mem. Cogn. 28, 108–124. doi:10.3758/BF03211580

Blanchette, I., and Dunbar, K. (2001). Analogy use in naturalistic settings: theinfluence of audience, emotion, and goals. Mem. Cogn. 29, 730–735. doi:10.3758/BF03200475

Catrambone, R. (1997). “Reinvestigating the effects of surface and structural fea-tures on analogical access,” in Proceedings of the Nineteenth Annual Conferenceof the Cognitive Science Society, eds M. G. Shafto and P. Langley (Mahwah, NJ:Erlbaum), 90–95.

Christensen, B. T., and Schunn, C. D. (2007). The relationship of analogical dis-tance to analogical function and preinventive structure: the case of engineeringdesign. Mem. Cogn. 35, 29–38. doi: 10.3758/BF03195939

Crano, W. D., and Brewer, M. B. (2002). Principles and Methods of Social Research.Mahwah, NJ: Erlbaum.

Dunbar, K. (1995). “How scientists really reason: scientific reasoning in real-worldlaboratories,” in The Nature of Insight, eds R. J. Sternberg and J. E. Davidson(Cambridge, MA: MIT Press), 365–395.

Dunbar, K. (1997). “How scientists think: online creativity and conceptual changein science,” in Conceptual Structures and Processes: Emergence, Discovery andChange, eds T. B. Ward, S. M. Smith, and S. Vaid (Washington, DC: AmericanPsychological Association), 461–493.

Dunbar, K. (2001). “The analogical paradox: why analogy is so easy in nat-uralistic settings, yet so difficult in the psychology laboratory,” in Analogy:Perspectives from Cognitive Science, eds D. Gentner, K. J. Holyoak, and B.Kokinov (Cambridge, MA: MIT Press), 313–334.

Forbus, K., Gentner, D., Everett, J., and Wu, M. (1997). “Towards a computationalmodel of evaluating and using analogical inferences,” in Proceedings of the 19thAnnual Conference of the Cognitive Science Society (Hillsdale, NJ: Erlbaum).

Gentner, D. (1983). Structure mapping: a theoretical framework for analogy. Cogn.Psychol. 7, 155–170.

Gentner, D., and Landers, R. (1985). “Analogical reminding: a good match is hardto find,” in Proceedings of the International Conference on Systems, Man, andCybernetics (Tucson, AZ), 607–613.

Gentner, D., and Markman, A. B. (1997). Structure mapping in analogy andsimilarity. Cogn. Psychol. 25, 524–575.

Gentner, D., Ratterman, M., and Forbus, K. (1993). The roles of similarity intransfer: separating retrievability from inferential soundness. Am. Psychol. 52,45–56.

Gick, M. L., and Holyoak, K. J. (1980). Analogical problem solving. Cogn. Psychol.12, 306–355.

Gick, M. L., and Holyoak, K. J. (1983). Schema induction and analogical transfer.Cogn. Psychol. 15, 1–38.

Glynn, S. M. (2008). “Making science concepts meaningful to students: teach-ing with analogies,” in Four Decades of Research in Science Education:From Curriculum Development to Quality Improvement, eds S. Mikelskis-Seifert, U. Ringelband, and M. Brückmann (Münster: Waxmann),113–125.

Guth, W., Schmittberger, R., and Schwarze, B. (1982). An experimental analysis ofultimatum bargaining. J. Econ. Behav. Organ. 3:367.

Hofstadter, D. (2001). “Analogy as the core of cognition,” in Analogy: Perspectivesfrom Cognitive Science, eds D. Gentner, K. J. Holyoak, and B. Kokinov(Cambridge, MA: MIT Press), 499–538.

Holyoak, K. J., Gentner, D., and Kokinov, B. N. (2001). “Introduction: the placeof analogy in cognition,” in Analogy: Perspectives from Cognitive Science, eds D.Gentner, K. J. Holyoak, and B. Kokinov (Cambridge, MA: MIT Press), 1–19.

Holyoak, K. J., and Koh, K. (1987). Surface and structural similarity in analogicaltransfer. Mem. Cogn. 15, 323–340.

Holyoak, K. J., and Thagard, P. (1995). Mental Leaps: Analogy in Creative Thought.Cambridge, MA: MIT Press.

Holyoak, K. J., and Thagard, P. (1997). The analogical mind. Am. Psychol. 52, 35–44.Hummel, J. E., and Holyoak, K. J. (1997). Distributed representations of

structure: a theory of analogical access and mapping. Psychol. Rev. 104,427–466.

Krawczyk, D. C., Holyoak, K. J., and Hummel, J. E. (2004). Structural constraintsand object similarity in analogical mapping and inference. Thinking Reasoning,10, 85–104. doi: 10.1080/13546780342000043

Krawczyk, D. C., Holyoak, K. J., and Hummel, J. E. (2005). The one-to-oneconstraint in analogical mapping and inference. Cogn. Sci. 29, 29–38. doi:10.1207/s15516709cog0000_27

Polya, G. (1957). How to Solve it. Princeton, NJ: Princeton University Press.Richland, L. E., Zur, O., and Holyoak, K. (2007). Cognitive supports for analo-

gies in the mathematics classroom. Science 316, 1128–1129. doi: 10.1126/sci-ence.1142103

Saner, L., and Schunn, C. (1999). “Analogies out of the blue: when history seems toretell itself,” in Proceedings of the 21st Annual Conference of the Cognitive ScienceSociety (Mahwah, NJ: Erlbaum).

Spellman, B. A., and Holyoak, K. J. (1996). Pragmatics in analogical mapping. Cogn.Psychol. 31, 307–346.

Wharton, C. M., Holyoak, K. J., Downing, P. E., Lange, T. E., Wickens, T. D.,and Melz, E. R. (1994). Below the surface: analogical similarity and retrievalcompetition in reminding. Cogn. Psychol. 26, 64–101.


Received: 30 May 2014; accepted: 03 November 2014; published online: 26 November2014.Citation: Kretz DR and Krawczyk DC (2014) Expert analogy use in a naturalisticsetting. Front. Psychol. 5:1333. doi: 10.3389/fpsyg.2014.01333This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Kretz and Krawczyk. This is an open-access article dis-tributed under the terms of the Creative Commons Attribution License (CC BY).The use, distribution or reproduction in other forums is permitted, providedthe original author(s) or licensor are credited and that the original publica-tion in this journal is cited, in accordance with accepted academic practice. Nouse, distribution or reproduction is permitted which does not comply with theseterms.



















GENERAL COMMENTARYpublished: 22 December 2014doi: 10.3389/fpsyg.2014.01487

Explaining the abundance of distant analogies innaturalistic observations of expertsMáximo Trench*

Psychology Department, Centro Regional Universitario Bariloche, National University of Comahue, Bariloche, Argentina*Correspondence: [email protected]

Edited by:


Reviewed by:

Daniel C. Krawczyk, The University of Texas at Dallas and University of Texas Southwestern Medical Center, USA

Keywords: analogy, naturalistic settings, experts, surface similarty, retrieval

A commentary on

Expert analogy use in a naturalistic

by Kretz, D., and Krawczyk, D. C. (2014).Front. Psychol. 5:1333. doi: 10.3389/fpsyg.2014.01333

Analogical reasoning is a landmark ofhuman cognition. Based on the realiza-tion that the elements of two situations areorganized by similar systems of relations,analogical inferences allow the transfer ofknowledge structures from a better-knownsituation (the base analog) to a target situ-ation that is relatively less understood (thetarget analog).

Experimental research has demon-strated that the retrieval of base analogsfrom long term memory in responseto the proceesing of a target analog isinfrequent in the lack of semantic simi-larities between both situations (Gick andHolyoak, 1980; Keane, 1987; Gentner et al.,1993; Trench and Minervino, 2014). Withthe turn of the century, several natural-istic observations of experts working intheir domains of expertise yielded a morecomplex picture. While molecular biolo-gists (Dunbar, 1997) and psychologists(Saner and Schunn, 1999) still exhib-ited mostly within-domain analogizing,the observation of journalists and politi-cians (Blanchette and Dunbar, 2001),teachers (Richland et al., 2004), man-agers (Bearman et al., 2007) and designengineers (Christensen and Schunn,2007) showed a more frequent use oflong-distance analogies. The naturalis-tic study by Kretz and Krawczyk (2014)on the use of analogies by economists also

demonstrates an abundance of distantanalogies in the sevice of an impres-sive variety of communicative purposes,most of which were not evident in priorresearch. These goals included the gen-eration of concrete source examples ofmore general target concepts, the forma-tion of visual images of source concepts,the addition of colorful speech, the inclu-sion of a target into a source concept, orthe differentiation between source and tar-get concepts. With these results in mind,the time is ripe to assert that the nat-uralistic observation of experts shows amore flexible use of analogical sourcesthan is predicted by experimental stud-ies on analogical transfer, and simulatedby dominant computer models of ana-logical retrieval (e.g., MAC/FAC, Forbuset al., 1995; LISA, Hummel and Holyoak,1997). How, then, to explain this analog-ical abundance? In trying to account forthe contrasting results of the experimentaland the naturalistic traditions, the defaultexplanations revolve around the expertiseof the analogizers and the psychologicalconstraints of the target tasks. I will arguethat although both factors are likely tobear some responsibility for this empiricalinconsistency, there are reasons to expect aheavier weight of the latter.

THE EXPERTISE OF THE ANALOGIZERSShortly after having documented thatjournalists and politicians generatedmostly distant analogies when argu-ing for (or against) the referendum onthe independence of Quebec, Blanchetteand Dunbar (2000) obtained an in vitroreplication of this result with novice par-ticipants generating their own analogies

for another realistic political topic: thezero-deficit strategy for controlling pub-lic debt. The authors concluded that theirprior results were due to the fact thatthe analogizers were generating theirown analogies for a realistic situation,rather than to their expertise in the targetissue. Trench et al. (2009a) provided sup-port for this interpretation by replicatingBlanchette and Dunbar’s results with 10different target topics. In the same vein,Bearman et al. (2007) failed to observedifferences in the analogies proposed bynovices and experts solving managementproblems. Rather than based on broadexpertise differences across-participants,it seems that the ease of generating dis-tant analogies depends on the goals ofthe analogizer and on the extent to whichshe understands the target analog at stake.When the analogizer comprehends thetarget analog better than her intendedaudience, as in communicative situationssuch as explaining a procedure to students(Richland et al., 2004) or selling politi-cal ideas to the population (Blanchetteand Dunbar, 2000, 2001; Trench et al.,2009a), both experts and novices easilygenerate distant analogies. But when thetarget analog is insufficiently understood,as when Dunbar’s (1997) expert molecularbiologists or Gick and Holyoak’s (1980)novice participants are attempting to solvea problem, distant analogies are seldomgenerated.

PSYCHOLOGICAL CONSTRAINTS OFTHE TARGET TASKAnother explanation for the frequentuse of distant analogies in naturalis-tic studies might arise from comparing

www.frontiersin.org December 2014 | Volume 5 | Article 1487 |

setting

116









http://journal.frontiersin.org/Journal/10.3389/fpsyg.2014.01333/abstract

http://journal.frontiersin.org/Journal/10.3389/fpsyg.2014.01333/abstract



Trench Expertise effects on analogy generation

the psychological constraints of natural-istic analogy generation against those ofclassical experimental studies. The stan-dard experimental procedure comprisesan encoding phase, during which partici-pants learn the base analogs, and a transferphase, where experimenters present par-ticipants with either a semantically closeor a semanically distant target situation,and assess whether its processing elic-its the spontaneous retrieval of the baseanalog. Based on this procedure, differ-ences accross conditions were typicallytaken to demonstrate the centrality of sur-face similarities during analogical trans-fer. In contrast with this highly controlledenvironment—in which transfer can onlyoriginate in the retrieval of the criti-cal base analog from memory—in nat-uralistic settings participants are free togenerate analogies by means of retriev-ing their own base analogs (Blanchetteand Dunbar, 2000; Hofstadter and Sander,2013), identifying conceptual metaphors(Minervino et al., 2009), stumbling acrosssuitable analogs in the external environ-ment (Christensen and Schunn, 2005)or fabricating novel base analogs eitherby generating extreme cases out of thetarget analog (Clement, 1988), or byreinstantiating the relational structure ofthe target with a new set of elements(Olguín et al., 2013). Upon generatingcandidate analogies via any combinationof such mechanisms, the proportion ofclose vs. distant analogies that peopleproduce in naturalistic settings can alsoreflect a conscious editing of one typeof analogies in favor of the other type,depending on the purpose of the reasoner(Trench et al., 2009b).

The fact that the core componentsof our retrieval mechanisms are invari-ably set to favor semantically close baseanalogs (Gentner et al., 1993; Trench andMinervino, 2014) suggests that the above-mentioned generative mechanisms couldpossibly account for the frequency andthe diversity of the analogies generatedby experts. Future studies, both natural-istic and experimental, are required tounderstand how these overlooked anal-ogy generation methods interact withthe variety of goals that realistic anal-ogy generation can pursue, as eloquentlyrevealed by Kretz and Krawczyk’s (2014)

detailed analysis of the analogies producedby expert economists.

ACKNOWLEDGMENTThis work was supported by grants PICT2352 and PICT 2650 from the NationalAgency for Scientific and TechnicalResearch of Argentina (ANPCyT),grant C108 from National Universityof Comahue and grant PIB013 fromUniversidad Abierta Interamericana.

REFERENCESBearman, C. R., Ball, L. J., and Ormerod, T. C.

(2007). The structure and function of sponta-neous analogizing in domain-based problemsolving. Think. Reason. 13, 273–294. doi:10.1080/13546780600989686

Blanchette, I., and Dunbar, K. (2000). How analogiesare generated: the roles of structural and super-ficial similarity. Mem. Cognit. 28, 108–124. doi:10.3758/BF03211580

Blanchette, I., and Dunbar, K. (2001). Analogy usein naturalistic settings: the influence of audience,emotion, and goals. Mem. Cogn. 29, 730–735. doi:10.3758/BF03200475

Christensen, B. T., and Schunn, C. D. (2005).Spontaneous access and analogical incuba-tion effects. Creat. Res. J. 17, 207–220. doi:10.1080/10400419.2005.9651480

Christensen, B. T., and Schunn, C. D. (2007). Therelationship of analogical distance to analogicalfunction and pre-inventive structure: the case ofengineering design. Mem. Cogn. 35, 29–38. doi:10.3758/BF03195939

Clement, J. (1988). Observed methods for gen-erating analogies in scientific problem solving.Cogn. Sci. 12, 563–586. doi: 10.1207/s15516709cog1204_3

Dunbar, K. (1997). “How scientists think: onlinecreativity and conceptual change in science,” inCreative Thought: An Investigation on ConceptualStructures and Processes, eds T. B. Ward, S. M.Smith, and S. Vaid (Washington, DC: APA Press),461–493.

Forbus, K., Gentner, D., and Law, K. (1995).MAC/FAC: a model of similarity-basedretrieval. Cogn. Sci. 19, 141–204. doi:10.1207/s15516709cog1902_1

Gentner, D., Rattermann, M. J., and Forbus, K.D. (1993). The roles of similarity in trans-fer: Separating retrievability from inferentialsoundness. Cogn. Psychol. 25, 431–467. doi:10.1006/cogp.1993.1013

Gick, M., and Holyoak, K. J. (1980). Analogicalproblem solving. Cogn. Psychol. 12, 306–355 doi:10.1016/0010-0285(80)90013-4

Hofstadter, D., and Sander, E. (2013). Surfaces andEssences: Analogy as the Fuel and Fire of Thinking.New York, NY: Basic Books.

Hummel, J. E., and Holyoak, K. J. (1997). Distributedrepresentations of structure: a theory of analogi-cal access and mapping. Psychol. Rev. 104, 427–466.doi: 10.1037/0033-295X.104.3.427

Keane, M. T. (1987). On retrieving analogues whensolving problems. Q. J. Exp. Psychol. 39, 29–41. doi:10.1080/02724988743000015

Kretz, D., and Krawczyk, D. C. (2014). Expert analogyuse in a naturalistic setting. Front. Psychol. 5:1333.doi: 10.3389/fpsyg.2014.01333

Minervino, R., López Pell, A., Oberholzer, N., andTrench, M. (2009). “A continuist approach topromoting creativity: generating novel metaphor-ical expressions through varying conceptualmetaphors,” in New Frontiers in Analogy Research,eds B. Kokinov, D. Gentner, and K. Holyoak (Sofia:NBU Press), 330–337.

Olguín, V., Trench, M., and Minervino, R. (2013).“Promoting interdomain analogical transfer: whencreating a problem helps solving a problem,” inPoster Presented at Analogy 2013 (Dijon).

Richland, L. E., Holyoak, K. J., and Stigler, J.W. (2004). Analogy use in eighth-grade math-ematics classrooms. Cogn. Instr. 22, 37–60. doi:10.1207/s1532690Xci2201_2

Saner, L., and Schunn, C. D. (1999). “Analogies outof the blue: when history seems to retell itself,”in Proceedings of the 21st Annual Conferenceof the Cognitive Science Society, eds M. Hahnand S. Stoness. (Mahwah, NJ: Erlbaum),619–624.

Trench, M., and Minervino, R. (2014). The role of sur-face similarity in analogical retrieval: Bridging thegap between the naturalistic and the experimen-tal traditions. Cogn. Sci. doi: 10.1111/cogs.12201.[Epub ahead of print].

Trench, M., Oberholzer, N., Adrover, F., andMinervino, R. (2009a). La eficacia del paradigmade producción para promover la recuperación deanálogos base interdominio. Psykhé 18, 39–48. doi:10.4067/S0718-22282009000100004

Trench, M., Oberholzer, N., and Minervino, R.(2009b). “Dissolving the analogical paradox:retrieval under a production paradigm is highlyconstrained by superficial similarity,” in NewFrontiers in Analogy Research, eds B. Kokinov,D. Gentner, and K. Holyoak (Sofia: NBU),443–452.

Conflict of Interest Statement: The author declaresthat the research was conducted in the absence of anycommercial or financial relationships that could beconstrued as a potential conflict of interest.

Received: 10 November 2014; paper pending pub-lished: 02 December 2014; accepted: 03 December 2014;published online: 22 December 2014.Citation: Trench M (2014) Explaining the abundance ofdistant analogies in naturalistic observations of experts.Front. Psychol. 5:1487. doi: 10.3389/fpsyg.2014.01487This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Trench. This is an open-access arti-cle distributed under the terms of the Creative CommonsAttribution License (CC BY). The use, distribution orreproduction in other forums is permitted, provided theoriginal author(s) or licensor are credited and that theoriginal publication in this journal is cited, in accor-dance with accepted academic practice. No use, distribu-tion or reproduction is permitted which does not complywith these terms.














Can taking the perspective of an expert debias humandecisions?The case of risky and delayed gainsMichał Białek*† and Przemysław Sawicki †

Department of Economic Psychology, Centre for Economic Psychology and Decision Sciences, Kozminski University, Warsaw, Poland

Edited by:


Reviewed by:

Gordon Pennycook, University ofWaterloo, CanadaFrancesca Chiesi, Università diFirenze, Italy

*Correspondence:

Michał Białek, Department ofEconomic Psychology, Centre forEconomic Psychology and DecisionSciences, Kozminski University,Jagiellonska 59, Warszawa 03-301,Polande-mail: [email protected]†Michał Białek and PrzemysławSawicki have contributed equally tothis work.

In several previously reported studies, participants increased their normative correctnessafter being instructed to think hypothetically, specifically taking the perspective of anexpert or researcher (Beatty and Thompson, 2012; Morsanyi and Handley, 2012). Thegoal of this paper was to investigate how this manipulation affects risky or delayedpayoffs. In two studies, participants (n = 193) were tested online (in exchange for money)using the adjusting procedure. Individuals produced certain/immediate equivalents forrisky/delayed gains. Participants in the control group were solving the problem from theirown perspective, while participants in the experimental group were asked to imagine“what would a reliable and honest advisor advise them to do.” Study 1 showed thatwhen taking the perspective of an expert, participants in experimental group becamemore risk aversive compared to participants in the control group. Additionally, their certainequivalents diverged from the expected value to a greater extent.The results obtained fromthe experimental group in Study 2 suggest that participants became less impulsive, whichmeans they tried to inhibit their preferences. This favors the explanation, which suggeststhat the perspective shift forced individuals to override their intuitions with the social norms.Individuals expect to be blamed for impatience or risk taking thus expected an expert toadvise them to be more patient and risk aversive.

Keywords: risk, intertemporal choices, expert, dual-process theory, decision-making

INTRODUCTIONHuman life presents continuous choices. In this text, we will focuson choices made in risk conditions (sticking to a permanent postvs. starting your own business) and intertemporal ones (buyingan iPad vs. saving money). Studies have shown that people makemistakes in both areas, which can lead to serious social problems,such as gambling or obesity.

Overwhelming media advertisements and marketing strengthenimpulsive behaviors in the modern society. This results foran example in obesity and financial debts. Nowadays, peopleare often facing artificial risky problems (stock market, infla-tion) for which they are not prepared (for extended review, seeTodd and Gigerenzer, 2012). This suggests a need to provide sup-port to people so they can make more rational decisions, especiallythose that are intertemporal and risky.

One of the recently introduced methods to support thinking,specifically reasoning, is hypothetical thinking. People are askedto assess a problem from the perspective of an expert or researcher(Beatty and Thompson, 2012; Morsanyi and Handley, 2012). Thisusually results in increased normative correctness of their mentalprocessing. Our aim is to test this method in a new field of cog-nition, that is, decision-making under risk and delay. We hope tovalidate the method of taking the perspective of an expert as anefficient debiasing method.

DECISIONS UNDER RISKUncertainty about the future is an inherent part of human exis-tence. While there are events we can be sure of and others that

are impossible to predict, most of us have to deal with probabilis-tic situations. Studies on choices in risky conditions show thatpeople have difficulty understanding information about probabil-ities. In a classic study, Tversky and Kahneman (1971) showedthat when people assess the probability of events, they tendto ignore base rate information and instead rely on the socialstereotype.

One of the main assumptions of prospect theory is that inrisky situations, people underestimate moderate and large proba-bilities but overrate rare events (Kahneman and Tversky, 1979).Failure to understand the rules of probability theory as wellas the fact that people overrate small probabilities are possi-ble causes for the high percentage of people participating indifferent types of gambling. Despite the unfavorable profit-to-risk ratio, studies have shown that 82% of adult Americans(Welte et al., 2002), 72% of Canadians (Azmier and Clements,2001), and 68% of adult British citizens (Wardle, 2007) admitto gambling. Even part of the stock market investors treatinvesting as a substitute of gambling. (Markiewicz and Weber,2013).

Biased perception of randomness is a challenge in the health-care domain. For example, in the medical context, there is thequestion of informing patients about the probability of variousdiseases. Much empirical evidence has shown that people haveserious problems estimating small probabilities. In particular,people are insensitive to changes in the magnitude of these prob-abilities (Kunreuther et al., 2001; Siegrist et al., 2008; Tyszka andSawicki, 2011).













Białek and Sawicki Debiasing decisions by taking the expert perspective

INTERTEMPORAL CHOICESIntertemporal choices present people with different challenges. Ineveryday life and in politics or economic affairs, some decisionsare based on choosing between payments that occur at a timedifferent from the time when the decision is made. For example,deciding whether to eat fast food immediately or wait for a bal-anced meal is based on the same psychological mechanisms asdeciding whether to spend your profit immediately or invest it.The issue of intertemporal choices is also examined in terms ofself-control, for example, not succumbing to temptation. In theclassic research commonly known as the marshmallow test, WalterMischel offered a four-year-old child a sweet and said if the childdecided not to eat it, he or she would soon get two marshmallows.The experimenter then left the room, leaving the child alone withthe temptation. If the child did not wait, he or she got only onemarshmallow instead of two. The results showed that people whohad greater self-control when they were children, scored higher onthe SAT test several years later, exhibited fewer behavioral prob-lems, coped better with stress, and were more focused and attentive(Mischel et al., 1989).

The inability to defer gratification may also lead to serious socialproblems, such as obesity. Obesity is estimated to be the seventhleading cause of mortality in the world (Ezzati et al., 2002). In2007–2008, 68% of adults in the US were overweight, and 33.8%of them were obese (Flegal et al., 2010).

SOCIAL NORMS AND NORMATIVE MODELSFor many cognitive processes there is a normative model, whichstates what is correct (e.g., logic for reasoning). For risk taking,multiplication of gain and probability of its occurrence expectedvalue (EV) is used as a normative model. Those who expect morefor a lottery than its EV are overly risk seeking, and those whoexpect less thank the EV are risk averse. The intertemporal choicesdo not have a normative model, but because of changes in oursocieties and extended lifespan, patience (to some degree) is seenas rational.

Social norms also regulate the behavior. Typically, patienceand the ability to avoid acting impulsively are virtues (Haidt andJoseph, 2004). Children are rewarded when they show the abilityto postpone reward.

Rational, according to the normative model of risk taking,would be to take a well calculated risk. It is unknown whetherthere are any consistent social norms regarding risk-taking, butexperiencing a loss because of risk-taking (action) is blamed morethan missing a chance to profit (omission, Ritov and Baron, 1990,1995). This happens because people expect to be blamed when tak-ing the risk, and risk avoidance can be seen as a socially acceptedbehavior.

METHODS IMPROVING DECISION MAKINGSome studies have focused on debiasing individuals in their con-clusions and decisions. These studies introduced different typesof instruction or additional information to help people overridetheir initial, biased intuitions. There are two, usually implicit,assumptions behind these manipulations.

First group of researchers tries to inform people about thenormative models and procedures (presenting people with the

concept of validity, EV or base rates). They assume that peopleare making mistakes because they lack the appropriate knowledgeor intuitions regarding the field of probability or logic (or mind-ware, as called by Stanovich, 2009b). In this view, an efficientmethod of debiasing would be a request to rely on a specific, for-mal procedure, e.g., “being presented with the concept of logicalvalidity, please try to assess following conclusions according totheir validity” (Evans et al., 1993).

The other group of researchers believes that biased thinking isnot a result of the lack of appropriate knowledge but of cognitivemiserliness (Fiske and Taylor, 1991), which means that individualsare making biased decision because of lack of available cognitiveresources and/or motivation to use reflexive processing. Whenmotivated and having enough time, individuals show less biaseddecision-making. In this view, an efficient method of debiasing isan instruction that relates to the procedure, supporting a deeperand reflexive thinking. An example of such instruction would be,“please try to override your initial beliefs and focus on the logi-cal structure of presented problems,” like used by Moutier et al.(2002).

Both approaches did not produce any satisfying and consis-tent increase in normative correctness of decisions. Despite theconsensus in the literature that debiasing requires decoupling ofthe intuitions with effortful processing (Croskerry et al., 2013),it seems that people are having troubles willingly override theirinitial beliefs, even when instructed to do so and when they aremotivated and have appropriate knowledge.

In contemporary literature, we can also find other meth-ods of improving individual’s cognitive processing. Hypotheticalthinking can increase the normative correctness of decisions byincreasing chances of using effortful, rule based processing (calledType 2, Evans and Over, 1996; Evans, 2007) but also inhibit-ing intuitive and heuristic responses (Type 1 processing). Forexample, Loewenstein et al. (2001) instructed participants to imag-ine the consequences of both presented alternatives, and thusdebiased people from the vividness effect. Lord et al. (1984)instructed their participants to imagine the opposite when con-sidering social dilemmas (e.g., capital punishment). Thanksto this strategy, they improved the objectivity of their judg-ments. Baron (2008) proposed open-minded thinking to helpoverride simple heuristic-cued intuitions. Trippas (unpublisheddoctoral dissertation) showed that cognitive style, understood aswillingness to engage in Type 2 processing rather than cogni-tive ability, influences accuracy of reasoning. This suggests thathuman cognition can be improved by encouraging people to thinkhypothetically (Type 2 processing).

Considering a problem from an expert’s perspective is quitea natural instruction. People typically perceive experts as unbi-ased and reliable sources of information. When taking an expert’sperspective, intuitions and emotions should play a minor role,and thanks to this manipulation, reflexive processing (1) shouldbe used more often compared to standard cases and (2) shouldoverride the internal conflict with intuitions.

In some studies, the instructing participants to take theexperts perspective effectively encouraged the use of Type 2processing, leading to less biased human performance (seeGreenhoot et al., 2004; Thompson et al., 2005; Ferreira et al., 2006;





Beatty and Thompson, 2012). However, when testing children, theeffect of the instruction was ambiguous. The instruction, “pleaseanswer the questions taking the perspective of a perfectly logicaland rational person (pp. 328),” as used in the study of Chiesi et al.(2011), seemed to work only for high cognitive ability individuals.This is possibly a result of the lack of learned rules and procedures,which could be applied to the task in the reported study or lack ofcognitive resources. We can assume that the rules of thinking, justlike driving a car, require much more cognitive resources whenthey are just learned while requiring fewer resources with greaterpractice (Stanovich, 2009a).

QUESTIONS AND HYPOTHESISAs stated in the introduction section, people make many mistakeswhen dealing with risk and delay, what results in many social prob-lems, like debts, gambling, obesity, and many more. The presentedliterature suggests that instructing people to reflect on a problemfrom the perspective of an expert can significantly improve theperformance. We expect our participants to produce less biaseddecisions in the field of risk and delay management.

The performance in risky condition is expected to be moreconsistent with the behavior predicted by EV normative model.This would provide an evidence of less biased risk assessment.

The performance in intertemporal choices is expected to changedirection toward bigger patience. The ability to focus on biggerbut more delayed goals seems to be more adaptive in mod-ern society than is impulsivity; thus, our manipulation shouldboost patience. This assumption does not follow from any nor-mative model, as in the case of risk; instead, the general socialnorm supports patient behavior rather than short-term orientedbehavior (Haidt and Joseph, 2004). Hence, we expect partici-pants to follow the social norm of impulsivity avoidance andproduce lower discounting strength in the experimental condi-tion in which participants are instructed to take the perspective ofan expert.

STUDY 1We investigated the effect of forced hypothetical thinking onchoices under risk. Situations, such as considering how muchone is willing to invest to get a higher but uncertain return,are everyday problems, but the form of presenting the problemis relatively artificial, and people are evolutionally not preparedto deal with such problems. In other words, people can intu-itively deal with risk, but the way they are presented increasesthe chance of making a biased decision (Gigerenzer and Selten,2002).

By improving these types of judgment, we could enhance peo-ple’s entrepreneurship abilities as well as prevent risky, maladaptivebehaviors, such as gambling or smoking. The hypothetical think-ing can possibly increase the use of other competing intuitions orrule-based processing. Both should change the risk taking decisionin a more normative manner and help people accurately estimatethe EV of potential gains.

MATERIALS AND METHODSThe certain equivalent of a potential gain of $5000 with a prob-ability of 90, 70 and 30% was computed for every participant.

The experimental manipulation was an instruction, asking par-ticipants to consider the problem from one of two perspectives,experts‘ or own. Our manipulation should increase abstract hypo-thetical thinking by taking the perspective of an expert1. Throughthis manipulation, we expected to debias human reasoning anddiscover the mindware responsible for risk management. Thefollowing is an example of tasks used:

Imagine you have received a $5000 reward, which will be paid to you witha 70% chance. An investment fund is willing to rebuy your reward witha certain payment. Imagine what an honest and rational expert wouldadvise you to do – accept or refuse the given offer.

ADJUSTING PROCEDUREThe research scheme used in all experiments was based on theadjusting method, which is most popular in the discountingresearch (Yi et al., 2006; Benzion et al., 2011; Odum, 2011).This procedure enables one to determine balance points, i.e.,payment methods where the person being tested was indiffer-ent to two given alternatives, for example, between receiving asum x immediately and a sum y after a period of time t (inresearch on risky choices, between receiving a sum x for cer-tain, and a sum y with a determined probability). Thanks tothe adjusting method, the overestimation of expected gains canbe prevented, which is common if individuals are asked directly,e.g., how much they would expect for having to wait for theirgain.

The research was conducted using a specially developed com-puter program, which enabled us to determine the balance points.The characteristic feature of the adjusting method is that onechoice alternative adjusts its value depending on previous deci-sions made by the person being tested. For example, during thetest on risky payments, two cards depicting sums were displayedon the screen, one was bigger but uncertain (the card on theright side of the screen), and the other was smaller and certain(the card on the left side of the screen). The sum on the cardon the right was fixed, while the card on the left changed itsvalue depending on the subsequent choices of the person beingtested.

The participant’s task was to choose between the two alterna-tives. With every question, the person being tested was to chooseone of the values (certain or uncertain). In the first step, the par-ticipant was given a choice of $2500 for sure (information onthe card on the left side) or $5000 with a 70% chance of win-ning (information on the card on the right side). After choosingthe risky option, in the following step, the certain sum increasedby half of its previous value. Hence, this time, the person beingtested could choose $3750 for sure or $5000 with a 70% chanceof winning. When the participant chose the risky option oncemore, the certain value increased by half of the previous valueand amounted to $4375 ($3750 + $625) in the subsequent step.However, if in the next step ($4375 for sure or $5000 with a70% chance of winning), the person being tested chose the cer-tain sum, its value decreased by half of the previous change (by

1We have also manipulated the ownership of the gain (deciding for myself or forsomebody else’s gains), but it produced neither main effect nor interaction; thus,we do not report this manipulation in more details.





$312). To sum up, the certain value was adjusted to reflect pre-vious choices of the person being tested. After making the sixthchoice, the program calculated the equivalent point for the riskyalternative.

PARTICIPANTSParticipants (n = 105) were citizens of the United States recruitedby specialized company in exchange for payment. They were ran-domly assigned to one of four experimental conditions. From thisgroup, 27 participants who answered illogically (e.g., they wantedto receive more for a 30% chance of winning compared to 70%chance) were removed from the database prior to any further anal-ysis. Mean age, gender distribution, and number of individuals ineach experimental condition are presented in Table 1.

RESULTSWhen a person is risk averse his subjective value of a lottery islower. Figure 1 presents a line that connects mean subjective val-ues of each lottery computed for the own/expert perspective andcompared to the EV model. We can see that individuals who tookthe perspective of an expert were more risk averse. Additionally,their result is further from the normative EV line compared to thecontrol group individuals who solved the problem from their ownperspective.

To measure the attitude towards risk, the area under the curvewas analyzed2 (Myerson et al., 2001). The surface of the areaunder the curve is the sum of all trapezes set by the next balancepoint values in relation to the ordinate and abscissa (Figure 2).By using this measure, one should first assign the values in therange (0, 1) to a scale of delay and the subjective value of thediscounted values. In the case of the scale of delay, values aresubsequently divided by the highest delays used. The scale ofsubjective values of the discounted sums is converted in a sim-ilar way. Next, we calculate the sum of the fields of the createdtrapezes using the formula (a2–a1) [(b1 + b2)/2], where a1 anda2 are consecutive delays while b1 and b2 represent consecutivesubjective values of gains. The area under the empirical dis-counting curve is therefore the sum of all trapezes. The smallerthe area under the curve the bigger is the risk aversion of anindividual.

Attitude toward risk was computed for each participant. Later,the parameters were compared across experimental groups. The

2This measure has two significant advantages compared to an alternative analysis,which is based on comparing discounting parameters. First, it reduces the level ofskewness compared to the analysis of the distribution of discounting parameters(k); second, it is neutral, i.e., it does not refer directly to any specific mathematicalformula.

Table 1 | Information about participants in Study 1.

Experimental condition N Ratio of females Mean age (SD)

Expert perspective 39 38.5 34.077 (9.94)

Control group 39 43.6 33.58 (8.58)

Overall 78 41.0 33.8 (9.24)

general linear model showed the main effect of perspective[F(1,78) = 7.634; p < 0.01; η2 = 0.091]3, and the participantsin the experimental group showed bigger risk aversion. The meanareas under the curve are shown in Figure 3. The bars representthe mean field under the curve, with higher value indicating morepositive attitude toward risk (lower risk aversion).

DISCUSSIONWe found that taking the perspective of an expert significantlyinfluenced the risky choices. Individuals in the experimental con-dition were more risk aversive. By comparing the choices in thestudy to the line representing the EV, we can conclude that the deci-sions made in the experimental condition were less normativelycorrect.

This result is contrary to our expectations because previouslyreported studies in other domains suggested that the perspectivemanipulation increases the normative correctness of decisions.

There are two possible explanations. First, the concept of EV isnot known to participants or the social norm for risky decisions(risk avoidance) was made less salient thanks to the perspectiveshift. The first explanation seems less probable, as already childrenintuitively compare lotteries by multiplying gains and probabilities(Schlottmann, 2001). But if individuals would follow the socialnorm would expect to be blamed for unsuccessful risk taking andthus avoid doing so to greater extent.

Morsanyi and Handley (2012) showed that the use of Type 2processing does not guarantee the correctness of thinking. Par-ticipants instructed to think from the perspective of a rationalperson focused even more on stereotypes instead of base rateswhen solving a lawyer’s task (Kahneman and Tversky, 1973).

In a study 1000 people were tested. Among the participants therewere 5 engineers and 995 lawyers. Jack is a randomly chosen par-ticipant of this study. Jack is 36 years old. He is not married and issomewhat introverted. He likes to spend his free time reading sciencefiction and writing computer programs.

Most of individuals endorsed the conclusion that Jack is anengineer. They showed more interest in social stereotype thanin the probability of occurrence of a specific event and answeredagainst the odds. This case is especially interesting for understand-ing the naïve probabilities that humans calculate. Here, taking arational perspective (Type 2 processing) exacerbated the neglectof base rates.

In other study, Pennycook et al. (2014) showed that the baserates are available at an intuitive level, so the increase of biasedresponses in referred Morsanyi and Handley study is a cor-rupted mindware case, where a social stereotype was judgedas a more reliable source of information compared to the baserates. Additionally, in the field of risky decision, the instruc-tion manipulation or enhancement of hypothetical thinking didnot consistently improve the performance (Weinstein and Klein,1995). The manipulation of instructions (hypothetical thinking)could possibly increase efficiency when one has the appropriatemindware: knows how to calculate the probability and how to

3The analysis on full population of participants showed also a significant main effectof the perspective manipulation [F(1,104) = 5.187; p < 0.05; η2 = 0.077]. By theelimination of irrational responses, we wanted to reduce a possible noise and reportmost reliable effect of the study.





FIGURE 1 | Mean certain equivalents for own/expert perspective compared to expected value model.

FIGURE 2 |The diagram shows four trapezes created as a result of

connecting equivalent points, with the risk axis by line segments. Boththe risk and subjective value of the sum are standardized from the rangefrom 0 to 1. Value 1 on the risk axis corresponds to a certain value, and onthe subjective value axis the non-discounted sum. Own elaboration basedon: Myerson et al. (2001).

use it in real-life social problems. If the mindware is missing, theperformance can be even worse (Chiesi et al., 2011).

STUDY 2The aim of the second study was to investigate the effect of forcedhypothetical thinking on intertemporal choices. Such situationsare everyday problems, where one has to think about how muchhe/she is willing to invest now to get a higher but delayed return.By improving that type of judgment, we could persuade peo-ple to improve health or prevent such maladaptive behavior asovereating.

This assumption does not follow from any normative model, asin the case of risk, but the general social norm supports patience

FIGURE 3 | Mean area under the curve for uncertain gains.

rather than short-term oriented actions (Haidt and Joseph, 2004).Thus, we expected participants to follow the social norm of impul-sivity avoidance and produce lower discounting strength in theexperimental condition of taking the perspective of an expert.

MATERIALS AND METHODSWe repeated the procedure of Study 1. The only difference was thatindividuals had to evaluate three delayed gains. The delay was setfor 1 month, 6 months, and 24 months. Once again, participantswere randomly assigned to one of two experimental conditions,own perspective or the perspective of an expert. Additionally, inStudy 2, we manipulated the ownership of the money (own moneyor someone else’s money), but this manipulation produced no





main effect and did not interact with the perspective manipulation;thus, it will not be reported in following analysis.

PARTICIPANTSParticipants (n = 130) were citizens of the United States recruitedby external services in exchange for payment. They were randomlyassigned to one of two experimental conditions. Details of thegroup are presented in Table 2. A small group of participants(n = 15) was excluded because of irrational responses (the samecriterion as in Study 1 was used).

RESULTSThe discounting curves that connect balance points for delayedgain of $5000 are presented in Figure 4, from which we can seethat when taking the perspective of an expert, people show lessimpulsivity compared to the control condition (own perspective).

General linear model analysis was used to test the influence ofpossession and perspective on discounting strength. Once again,a main effect of perspective has been found [F(1,114) = 4.168;p < 0.05; η2 = 0.036]4. The mean discounting strength ispresented in Figure 5.

DISCUSSIONTaking the perspective of an expert improves thinking by helpingindividuals overcome impulsivity. Participants taking an expert’sperspective were less likely to lose some of their money by receiv-ing it sooner and not having to wait compared to considering aproblem from their own perspective. We can conclude that indi-viduals’ mindware related to delay management is governed bya rule, where impulsivity is assessed as wrong, and the correctbehavior is patience. This belief is then in conflict with the intu-itive willingness to receive immediate rewards. This is consistentwith common observations that people are sometimes consciouslyaware of the internal conflict between their intuitions (impulsivity,Type 1 processing) and beliefs about correct response (patience,Type 2 processing).

Despite the lack of normative model, we can conclude thatthe patience in this specific task presented here is adaptive; thus,it should be evaluated positively. This internal conflict shouldemerge when the mindware response is made salient by takingthe perspective of an expert. De Neys (2014) stated that when theemotional arousal emerges after the conflict detection and an indi-vidual notices it, the heuristic response could be questioned. Thiscan decrease the impulsivity, as it is a heuristic response.

4As requested by reviewer, we tested gender as a factor in the GLM analysis andfound no main effect or interaction (both p > 0.5).

Table 2 | Information about participants in Study 2.

Experimental condition N Ratio of females Mean age (SD)

Experts perspective 62 35 34.4 (10.31)

Own perspective 53 57.9 37.5 (11.46)

Overall 115 46.15 35.91 (10.92)

FIGURE 4 | Discounting curves for delayed gains.

FIGURE 5 | Discounting strength for delayed gains.

GENERAL DISCUSSIONThe perspective shift changed human decisions. Participantsshowed bigger patience and risk aversion. The results in the areaof risk are not consistent with our expectations. The decisionsmade cannot be seen as more correct or rational. The intertempo-ral choices have shown an improvement, as the discounting rateshown by individuals decreased. This can help individuals over-come temptations and help them make long-term financial plans(savings, investments).

Possible explanation of observed behavior is that individu-als focused on social norms rather than on normative models.This would be consistent with findings of Chiesi et al. (2011)and Morsanyi and Handley (2012) who reported no consistentimprovement or even decrease in correctness of decisions underthe forced perspective shift. We discuss, that the experimen-tal manipulation made the social norm salient and people whotook experts’ perspective focused on the social norm to big-ger extent, than in typical, everyday decision. Because patienceand cautiousness are socially perceived as virtues we see a





change of decisions to match those. This assumption is consis-tent with the reported findings on the improvement of thinkingand reasoning presented in the introduction section. The socialnorm related to thinking promotes reflexive and logical think-ing, that is, the Type 2 processing. The mindware responsiblefor dealing with risk is still a main topic for research on riskand delay management, but social norms related to this topichave to be investigated more and incorporated in the design ofstudies.

Our data did not fully support presented conclusion, as we havenot tested the social norms of participants. We assume that thecultural norm should have some effect on most individuals. Theproverb “A bird in the hand is worth two in the bush,” is a goodexample of socially accepted risk avoidance. The hypothesis ofsocial norms made salient could be tested by comparing cultureswith big differences in the delay management or in the attitudetoward risk.

Despite possible difficulties, the idea of improving individu-als’ decision-making in the area of risk and delay seems to beworth effort. Individuals perform sub-optimally even in advanta-geous conditions (with all required information provided and notime pressure) and require to be supported. Modern society cre-ated an artificial environment (e.g., by marketing, commercials)in which people are misinformed or put under time pressure; thus,human decision-making needs to be supported to greater extent,particularly by hypothetical thinking with the focus on a specificinstruction.

ACKNOWLEDGMENTSThe current project was financed by the resources of PolishNational Science Centre (NCN) assigned by the decision no.2013/11/D/HS6/04604; http://www.nauka.gov.pl. The funders hadno role in study design, data collection and analysis, decision topublish, or preparation of the manuscript. Authors would like tothank to Prof. Tadeusz Tyszka and Dr. Łukasz Markiewicz for theircomments and advices.

REFERENCESAzmier, J. J., and Clements, M. (2001). Gambling in Canada. Calgary, AB: Canada

West Foundation.Baron, J. (2008). Thinking and Deciding. New York: Cambridge University Press.

doi: 10.1017/CBO9780511840265Beatty, E. L., and Thompson, V. A. (2012). Effects of perspective and belief on

analytic reasoning in a scientific reasoning task. Think. Reason. 18, 441–460. doi:10.1080/13546783.2012.687892

Benzion, U., Krahnen, J., and Shavit, T. (2011). Subjective evaluation of delayedrisky outcomes for buying and selling positions: the behavioral approach. Ann.Financ. 7, 247–265. doi: 10.1007/s10436-010-0172-4

Chiesi, F., Primi, C., and Morsanyi, K. (2011). Developmental changes in proba-bilistic reasoning: the role of cognitive capacity, instructions, thinking styles, andrelevant knowledge. Think. Reason. 17, 315–350. doi: 10.1080/13546783.2011.598401

Croskerry, P., Singhal, G., and Mamede, S. (2013). Cognitive debiasing 1: originsof bias and theory of debiasing. BMJ Qual. Saf. 22(Suppl. 2), ii58–ii64. doi:10.1136/bmjqs-2012-001712

De Neys, W. (2014). Conflict detection, dual processes, and logical intuitions:Some clarifications. Think. Reason. 20, 169–187. doi: 10.1080/13546783.2013.854725

Evans, J. St. B. T. (2007). Hypothetical Thinking: Dual Processes in Reasoning andJudgment. New York: Psychology Press.

Evans, J. St. B. T., Newstead, S. E., and Byrne, R. M. J. (eds). (1993). HumanReasoning: The Psychology of Deduction. Hove: Lawrence Erlbaum Associates.

Evans, J. St. B. T., and Over, D. (1996). Rationality and Reasoning. Hove: PsychologyPress.

Ezzati, M., Lopez, A. D., Rodgers, A., Vander Hoorn, S., and Mur-ray, C. J. (2002). Selected major risk factors and global and regionalburden of disease. Lancet 360, 1347–1360. doi: 10.1016/S0140-6736(02)11403-6

Ferreira, M. B., Garcia-Marques, L., Sherman, S. J., and Sherman, J. W.(2006). Automatic and controlled components of judgment and decisionmaking. J. Pers. Soc. Psychol. 91, 797–813. doi: 10.1037/0022-3514.91.5.797

Fiske, S. T., and Taylor, S. E. (1991). Social Cognition. Reading, MA: Addison-Wesley.

Flegal, K. M., Carroll, M. D., Ogden, C. L., and Curtin, L. R. (2010). Preva-lence and trends in obesity among US adults. JAMA 303, 235–241. doi:10.1001/jama.2009.2014

Gigerenzer, G., and Selten, R. (eds). (2002). Bounded Rationality: The AdaptiveToolbox. Cambridge, MA: MIT Press.

Greenhoot, A. F., Semb, G., Colombo, J., and Schreiber, T. (2004). Prior beliefs andmethodological concepts in scientific reasoning. Appl. Cogn. Psychol. 18, 203–221.doi: 10.1002/acp.959

Haidt, J., and Joseph, C. (2004). Intuitive ethics: how innately preparedintuitions generate culturally variable virtues. Daedalus 133, 55–66. doi:10.1162/0011526042365555

Kahneman, D., and Tversky, A. (1973). On the psychology of prediction. Psychol.Rev. 80, 237–251. doi: 10.1037/h0034747

Kahneman, D., and Tversky, A. (1979). Prospect theory: an analysis of decisionunder risk. J. Electrochem. Soc. 47, 263–291. doi: 10.2307/1914185

Kunreuther, H., Novemsky, N., and Kahneman, D. (2001). Making low prob-abilities useful. J. Risk Uncertain. 23, 103–120. doi: 10.1023/A:1011111601406

Loewenstein, G. F., Weber, E. U., Hsee, C. K., and Welch, N. (2001). Risk as feelings.Psychol. Bull. 127, 267–286. doi: 10.1037/0033-2909.127.2.267

Lord, C. G., Lepper, M. R., and Preston, E. (1984). Considering the opposite:a corrective strategy for social judgment. J. Pers. Soc. Psychol. 47, 1231. doi:10.1037//0022-3514.47.6.1231

Markiewicz, Ł., and Weber, E. U. (2013). DOSPERT’s gambling risk-taking propen-sity scale predicts excessive stock trading. J. Behav. Financ. 14, 65–78. doi:10.1080/15427560.2013.762000

Mischel, W., Shoda, Y., and Rodriguez, M. (1989). Delay of gratification in children.Science 244, 933–938. doi: 10.1126/science.2658056

Morsanyi, K., and Handley, S. J. (2012). “Does thinking make you biased? the case ofthe engineers and lawyers problem,” in Proceedings of the 34th Annual Conferenceof the Cognitive Science Society, eds N. Miyake, D. Peebles, and R. P. Cooper(Austin, TX: Cognitive Science Society), 2049–2054.

Moutier, S., Angeard, N., and Houde, O. (2002). Deductive reasoning and matching-bias inhibition training: evidence from a debiasing paradigm. Think. Reason. 8,205–224. doi: 10.1080/13546780244000033

Myerson, J., Green, L., and Warusawitharana, M. (2001). Area under thecurve as a measure of discounting. J. Exp. Anal. Behav. 76, 235–243. doi:10.1901/jeab.2001.76-235

Odum, A. L. (2011). Delay discounting: i’m a k, you’re a k. J. Exp. Anal. Behav. 96,427–439. doi: 10.1901/jeab.2011.96-423

Pennycook, G., Trippas, D., Handley, S. J., and Thompson, V. A. (2014). Base rates:both neglected and intuitive. J. Exp. Psychol. Learn. Mem. Cogn. 40, 544–554. doi:10.1037/a0034887

Ritov, I., and Baron, J. (1990). Reluctance to vaccinate: omission bias and ambiguity.J. Behav. Decis. Mak. 3, 263–277. doi: 10.1002/bdm.3960030404

Ritov, I., and Baron, J. (1995). Outcome knowledge, regret, and omission bias.Organ. Behav. Hum. Decis. Process. 64, 119–127. doi: 10.1006/obhd.1995.1094

Schlottmann, A. (2001). Children’s probability intuitions: understanding theexpected value of complex gambles. Child Dev. 72, 103–122. doi: 10.1111/1467-8624.00268

Siegrist, M., Cousin, M. E., and Keller, C. (2008). Risk communica-tion, prenatal screening, and prenatal diagnosis: the illusion of informeddecision-making. J. Risk. Res. 11, 87–97. doi: 10.1080/13669870701574015


http://www.nauka.gov.pl.




Stanovich, K. E. (2009a). What Intelligence Tests Miss: The Psychology of RationalThought. London: Yale University Press. doi: 10.5840/inquiryctnews201126216

Stanovich, K. E. (2009b). “Distinguishing the reflective, algorithmic, andautonomous minds: is it time for a tri-process theory,” in Two Minds:Dual Processes and Beyond, eds J. St. B. T. Evans and K. Frankish (Oxford:Oxford University Press), 55–88. doi: 10.1093/acprof:oso/9780199230167.003.0003

Thompson, V. A., Evans, J. St. B. T., and Handley, S. J. (2005). Persuadingand dissuading by conditional argument. J. Mem. Lang. 53, 238–257. doi:10.1016/j.jml.2005.03.001

Todd, P. M., and Gigerenzer, G. (2012). Ecological Rationality: Intelligence in theWorld. Oxford: Oxford University Press. doi: 10.1093/acprof:oso/9780195315448.001.0001

Tversky, A., and Kahneman, D. (1971). Belief in the law of small numbers. Psychol.Bull. 76, 105–110. doi: 10.1037/h0031322

Tyszka, T., and Sawicki, P. (2011). Affective and cognitive factors influenc-ing sensitivity to probabilistic information. Risk Anal. 31, 1832–1845. doi:10.1111/j.1539-6924.2011.01644.x

Wardle, H. (2007). British Gambling Prevalence Survey. The Stationery Office.London: National Centre for Social Research.

Weinstein, N. D., and Klein, W. M. (1995). Resistance of personal risk perceptionsto debiasing interventions. Health Psychol. 14, 132–140. doi: 10.1037/0278-6133.14.2.132

Welte, J. W., Barnes, G. M., Wieczorek, W. F., Tidwell, M.-C., and Parker, J. (2002).Gambling participation in the US—results from a national survey. J. Gambl. Stud.18, 313–337. doi: 10.1023/A:1021019915591

Yi, R., De La Piedad, X., and Bickel, W. K. (2006). The combined effectsof delay and probability in discounting. Behav. Processes 73, 149–155. doi:10.1016/j.beproc.2006.05.001


Received: 31 May 2014; accepted: 20 August 2014; published online: 04 September2014.Citation: Białek M and Sawicki P (2014) Can taking the perspective of an expertdebias human decisions? The case of risky and delayed gains. Front. Psychol. 5:989. doi:10.3389/fpsyg.2014.00989This article was submitted to Cognition, a section of the journal Frontiers in Psychology.Copyright © 2014 Białek and Sawicki. This is an open-access article distributed underthe terms of the Creative Commons Attribution License (CC BY). The use, distributionor reproduction in other forums is permitted, provided the original author(s) or licensorare credited and that the original publication in this journal is cited, in accordance withaccepted academic practice. No use, distribution or reproduction is permitted whichdoes not comply with these terms.







ORIGINAL RESEARCH ARTICLEpublished: 04 February 2014

doi: 10.3389/fpsyg.2014.00047

The geometry of expertiseMaría J. Leone1,2*, Diego Fernandez Slezak3, Guillermo A. Cecchi4 and Mariano Sigman1,2,5

1 Laboratorio de Neurociencia Integrativa, Departamento de Física, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Buenos Aires,Argentina

2 IFIBA, CONICET, Buenos Aires, Argentina3 Departamento de Computación, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Buenos Aires, Argentina4 Computational Biology Center, T.J. Watson Research Center, International Business Machines, Yorktown Heights, NY, USA5 Universidad Torcuato Di Tella, Buenos Aires, Argentina

Edited by:


Reviewed by:

Guillermo Campitelli, Edith CowanUniversity, AustraliaMichael H. Connors, MacquarieUniversity, Australia

*Correspondence:

María J. Leone, Laboratorio deNeurociencia Integrativa,Departamento de Física, FCEN UBAand IFIBA, CONICET; Pabellón 1,Ciudad Universitaria, 1428 BuenosAires, Argentinae-mail: [email protected]

Theories of expertise based on the acquisition of chunk and templates suggest adifferential geometric organization of perception between experts and novices. It is impliedthat expert representation is less anchored by spatial (Euclidean) proximity and mayinstead be dictated by the intrinsic relation in the structure and grammar of the specificdomain of expertise. Here we set out to examine this hypothesis. We used the domain ofchess which has been widely used as a tool to study human expertise. We reasoned thatthe movement of an opponent piece to a specific square constitutes an external cue andthe reaction of the player to this “perturbation” should reveal his internal representationof proximity. We hypothesized that novice players will tend to respond by moving a piecein closer squares than experts. Similarly, but now in terms of object representations,we hypothesized weak players will more likely focus on a specific piece and henceproduce sequence of actions repeating movements of the same piece. We capitalizedon a large corpus of data obtained from internet chess servers. Results showed that,relative to experts, weaker players tend to (1) produce consecutive moves in proximalboard locations, (2) move more often the same piece and (3) reduce the number ofremaining pieces more rapidly, most likely to decrease cognitive load and mental effort.These three principles might reflect the effect of expertise on human actions in complexsetups.

Keywords: chess expertise, object representation, chunks, spatial proximity, attentional control

INTRODUCTIONThe focus of attention can be directed by exogenous (bottom-up) and endogenous (top-down) cues (Pylyshyn, 2007; Richardet al., 2008). The region of visual space to which attention isdirected changes according to specific goals and tasks (Gilbert andSigman, 2007; Vinckier et al., 2007). A classic example is Yarbusgaze experiment (where subjects had to view a complex imageseveral times, each with a different instruction), he demonstratedthat the sequence of eye fixations changed drastically accordingto the question the observer was trying to respond about theimage (Yarbus, 1967; Tatler et al., 2010). Top-down control ofattention can act over a wide range of categories, including loca-tion but also objects, goals, features, context, time . . . (Duncan,1984; Chun and Jiang, 1998; Maunsell and Treue, 2006; Newet al., 2007). The ability to direct attention to specific objects andcategories changes with experience (Gilbert and Sigman, 2007)and, similarly, the ability to ignore salient cues requires inhibitionmechanisms which are trainable (Cepeda et al., 1998).

Chess has been one of the most widely studied modelsof expertise (de Groot, 1978; Schultetus and Charness, 1999;Reingold et al., 2001a,b; Campitelli and Gobet, 2008; Connorset al., 2011; Bilalic et al., 2012). Chess experts recognize andrecall chess positions accurately [chunk and templates theories(Chase and Simon, 1973; Gobet and Simon, 1996)], and developheuristics that allow them to focus and explore only a few “good

enough” moves (de Groot, 1978), substantially alleviating thesearch process.

As in other domains of perceptual expertise, chunk and tem-plates acquisition theories reflect geometrical differences in theorganization of perception between chess experts and novices:strong players recognize groups of pieces connected by functionalrelations as units (Gobet and Simon, 1996) and they also explorechess positions differently than novices [eye fixations are morecentered in relationships between pieces (Reingold et al., 2001a)].An implication of this theory is that expert representation is notanchored to the proximity between two pieces (Euclidean dis-tance) and may instead be dictated by the intrinsic relation in thestructure and grammar of the board. For instance, a bishop in onecorner of the board which works in concert with a knight on theother side of the board to jointly attack an opponent square maybe “functionally proximal” pieces in the mind of an expert, but“functionally distal” in the mind of a novice who does not recog-nize this relation. The same argument is true for other domainsof expertise, for instance an expert soccer goalkeeper may bindtogether (spatially) distant properties of the field (where is theball, where are the defenders and the attackers) which jointly maybuild an important cluster of features. Here we set out to examineexplicitly this conjecture. We focus on chess which has three prin-cipal advantages to solve our goal: (1) The degree of proficiencycan be quantified precisely with international systems of ratings

www.frontiersin.org February 2014 | Volume 5 | Article 47 | 126







http://www.frontiersin.org/people/u/63243







Leone et al. Expertise geometry

(Elo, 1978), (2) The spatial layout in chess is well delimited by asquare discrete grid of 64 locations, and (3) Capitalizing on chessinternet servers we can base our conclusions on massive sets ofdata.

We reasoned that the movement of an opponent piece to aspecific square constitutes an external attentional cue. The reac-tion of the player to this “perturbation” should reveal his internalrepresentation of proximity. Specifically, we hypothesized that anaive player will tend to respond in proximal squares. Instead,we hypothesized that expert player responses are less likely tobe governed by the spatial position of the opponent last move.Additionally, when this argument is expressed in terms of objectoriented attention, we hypothesized that a novice player will morelikely direct attention (concentrate) to a specific piece and henceproduce repeated sequence of moves with the same piece. Instead,expert representation is directed to a more sophisticated patternof pieces (chains of pawns, coordinated set of pieces working inconcert. . . ) and hence the sequence of moves should show fewerrepetitions. If weak players focus more on individual pieces, wehypothesized they should tend to reduce the number of objects tobe attended to avoid cognitive load. Finally, we examined somecore aspects of chess strategy hypothesizing the experts wouldplay more in-line with them than novices.

MATERIALS AND METHODSDATA ACQUISITIONAll games were downloaded from FICS (Free Internet Chess Server,http://www.freechess.org/), a free ICS-compatible server for play-ing chess games through Internet, with more than 300,000 reg-istered users. This constitutes a quite unique experimental setupproviding virtually infinite data (thousands of millions of moves).

Each registered user may be human or a computer player,and has associated a rating (Glicko rating, http://www.glicko.net/glicko.html) that indicates the chess skills strength of the player,represented by a number typically between 1000 and 3000 points.We defined two expertise levels: (a) strong players with ratinghigher than 1900 points and (b) weak players with rating between1000 and 1400 points.

A regular game of chess contains about 40 moves from eachplayer. A ply (plural plies) refers to one turn taken by a player.Hence, a chess game of 40 moves corresponds to 80 plies. We usethe term “next move” to refer to two consecutive actions by thesame player (white move 1: e4, white move 2: Nf3) and the term“next ply” to refer to consecutive actions by each player (ply 1:e4—white movement; ply 2: e5—black movement).

For each expertise group, we selected games from the FICSdatabase (played from 2005 to 2013) with at least 80 plies, playedbetween human players (of the same expertise group), with a totaltime budget of 180 s for each player and no increment. In order tomake further comparisons, we generated 35 sets for each expertisegroup (each with 5000 games). These sets were built by date (i.e.,set 1 contained mostly games from 2005 and set 35, from 2013).For each analysis we compared 35 values (one for each set) fromthe high rated expertise group with the 35 values from the lowrated group.

In order to extend the results to longer time budgets, we repli-cated the analysis with games of 300 and 900 s per player. Because

of the lower number of available games for these time budgets,we generated 35 datasets of 1500 games each (for each expertisegroup) for each of them. All other conditions were maintained.

MOVEMENT DISTANCE MEASUREPiece location coordinates were organized in an 8∗8 matrix (rep-resenting the 64 squares of the chess board). For each movement,we calculated two different measures (D1 and D2) to examine theproximity between successive actions in a game of chess. To dothis, we define the initial square of a movement as the location ofthe moving piece before the movement and the final square of amovement as the location of the moving piece after this action.

D1 corresponds to the difference between the initial square ofa piece movement and the final square of the previous move-ment by the opponent. In other words, this corresponds to thedifference in location between two consecutive plies of the game.This is the observable which can be more easily mapped to classicattentional cues experiments. We make the analogy that the loca-tion where the opponent drops a piece is an attentional cue andobserve the player’s onset of his response relative to this cue.

D2 corresponds to the difference between the final squares oftwo successive moves from the same player. This is measured asthe difference between the end-locations of ply(n + 2) and ply(n).

For each measure we calculated the signed difference insquares in the x and y axis. Positive values of the y-axis indicatethat the difference is shifted toward the upper side of the board.To do so we assumed the normal row conventions of chess (1indicating white’s first row and 8 black’s first row). A positive dif-ference in x-axis indicates a shift toward the left side and negativevalues a movement toward the right side of the board. Note thatthe differences in both axis take a range in the [−7, −6, . . . , 0,1, . . . , 7] values. A value of zero indicates that there was not shiftin that axis. For each measure (D1 and D2), we obtained a dis-tance matrix for white moves and other for black moves, and weadded them.

For each independent set (35 per expertise group) we cal-culated the 15∗15 matrices D1mat(i,R) and D2mat(i,R), where iranges from 1 to 35 and R can be high or low rated group.Probabilities for each entry of the matrices were calculated asfollow: (a)- for each movement (at least 40∗5000 –number ofmoves∗number of games-, white or black) we calculated the dis-tance measure, (b) we counted the amount of movements whichmatched with each matrix entry, (c) we divided the number ofmovements on each entry by the total number of movementson the distance matrix. For each matrix, the entries indicate theprobability to find successive plies (for D1) or moves (for D2) atthe corresponding distance (see Figure 1) where indices 1 . . . 15,respectively, code distances [−7 . . . 7].

Our main experimental question is to investigate whethercertain transitions (spatial differences between successive move-ments) differ between high and low rated players. To examinethis we performed two-sample t-tests comparing high vs. lowgroups. We performed an independent t-test for each entry of thematrix (a total of 225 tests), each comparing the 35 values mea-sured for the high and low expertise groups. From this analysiswe generated a matrix oft-values which encodes the difference inprobabilities for each entry between high and low rated players.

Frontiers in Psychology | Cognition February 2014 | Volume 5 | Article 47 | 127

http://www.freechess.org/

http://www.glicko.net/glicko.html

http://www.glicko.net/glicko.html





FIGURE 1 | Expertise level defines spatial effects on successive

movements. (A) Distance measurements. Distances between twosquares over an 8∗8 checkerboard can be measured subtracting thecoordinates of one location (white circle) to the coordinates of otherlocation (represented by the end of each arrow) on both x and yaxis. The square where the white circle is located corresponds tox = 6 and y = 6, and four alternative locations are illustrated withcolor arrows. All possible distances between two squares of this8 × 8 board are constrained to a 15∗15 square, where now thecoordinates of the white circle are (0, 0) and distance measuresrange from −7 to 7. The x axis shows the direction movement onnumber of columns (x < 0, left direction; x > 0, right direction) andthe y axis represents movements on number of the rows (y < 0,down direction; y > 0, up direction). For example, the distancebetween the end of the yellow arrow (which is at x = 5 and y = 7on the 8∗8 board) and the white circle is calculated as thedifference between the corresponding squares coordinates (�x = −1and �y = 1). (B) Probability distributions of movement distances. Weuse two observables to assess locality effects on chess playing: D1

and D2 (see Methods). Probabilities to make a movement close to

the previous one is higher at short distances, for both High andLow expertise levels. (C) Weak players made their movements closerto the previous one. We contrasted probability distributions for bothHigh and Low rated players on each entry of the 15∗15 distancesquare independently. t-value of each independent two-sample t-test(with p-value < 0.001, Bonferroni corrected for multiple comparisons)is color-coded. Positive (red) t-values indicate significantly higherprobabilities for high rated players and negative (blue) values, forweaker players. (D) Radial or Euclidean distances. Distances wereone-dimension collapsed and the difference between probabilities ofmaking movements corresponding to a distance square [P(High) –P(Low)] was plotted vs. each radial distance. High and low ratedgroups distributions were independently compared in each radialdistance [two-sample t-test on each variable (D1 and D2)]. Redasterisks (t > 5.3 for D1 and t > 5.6 for D2) indicates distances wereP(High) is significantly higher than P(Low); blue asterisks (t < −12.3 forD1 and t < −11 for D2), distances were P(Low) > P(High); in bothp < 0.001, Bonferroni corrected for multiple comparisons. Dotted linesindicate distance values where the significances changes from weakto strong players.





Positive t-values indicate that this transition is more likely forhigh than for low rated players. p-values were corrected with astrict Bonferroni criterion for multiple comparisons, consideringa difference as significant only if p < 0.001/225. For visualizationpurposes, t-values in comparisons that did not reach significancewere set to t = 0.

We then collapsed the matrices D1mat(i,R) to their radial dis-tance by the conventional formula r = (�x2 + �y2)1/2. Thisresulted on independent vectors of 34 dimensions for each i andR. Note that the possible differences in distance over the boardis less than 15∗15 since there is a lot of redundancy (for instancemoving the king one square to the left, to the right, up or down,all correspond to a radial distance of 1). As before, we convertedthese distributions in a vector of t-values which encode the differ-ence in probabilities for each index between high and low ratedplayers. Positive t-values indicate that this transition is more likelyfor high than for low rated players. p-values were corrected with astrict Bonferroni criterion for multiple comparisons, consideringa difference as significant only if p < 0.001/34.

PIECE REPETITIONThis analysis is shown only for white moves (analyzing blackmoves yielded identical results). We encoded for each white movewhether the piece moved was the same (1) or different (0) thanthe piece moved in the previous turn. The probability of movingthe same piece twice depends on the number of pieces remainingof the board. Thus, we calculated the repetition probability as afunction of the pieces remaining in the board independently forthe 35 sets of each expertise level. This yielded vectors PR(i,R,np)where i indicates the group (1–35), R the expertise level (high orlow) and np the number of pieces remaining on the board from 3to 16 (there were not sufficient positions with one or two piecesremaining on the board when considering the first 40 moves). Toquantify the main effects of expertise level and number of pieceson the repetition probability we made a Two-Way ANOVA testwith number of pieces and expertise level as independent factorsand their interactions. Then, we compared the distributions ofhigh and low rating independently for each number of pieces withindependent two-sample t-tests comparing high and low ratedplayers. p-values were corrected with a strict Bonferroni criterionfor multiple comparisons, considering a difference as significantonly if p < 0.001/16.

NUMBER OF PIECES REMAINING ON THE BOARDThis analysis is shown only for white moves (analyzing blackmoves yielded identical results). We calculated for each whitemove number (1–40) the number of white pieces remaining onthe board (1–16). We then averaged this value for each movenumber (1–40) across all the games in each set. This yieldedvectors NP(i,R,n) where i indicates the set (1–35), R the exper-tise level (high or low), and n the move number from 1 to 40.Note that this distribution as function of move number has to bemonotonous decreasing. This would not be true in bughouse orcrazy house variants of the games where a player can introduce acaptured piece back to the board. We assessed the main effectsof move number and expertise level with a Two-Way ANOVAtest with move number and expertise level as independent factors

and their interactions. Then, we compared the distributions ofhigh and low rating independently for each move number withindependent two-sample t-tests. p-values were corrected with astrict Bonferroni criterion for multiple comparisons, consideringa difference as significant only if p < 0.001/40.

BOARD DISTRIBUTIONThis analysis is shown only for white moves (analyzing blackmoves yielded identical results). We calculated for white moves5, 10, 20, 30, and 40 and for each piece the frequency of dis-tribution along all squares of the board. Since the remainingamount of pieces between varies with expertise levels and thenumber of types of pieces is not the same (8 pawns, 2 bishops, oneking,. . . ) we normalized the occurrences by the average numberof the piece type at each move number on each set. For exam-ple, the probability to have a knight on b1 in the move number1 (in a set of an expertise level) is 0.98 and for g1 is 0.93, butat that moment there were 2 knights over the board (on averagein that set), then the values for b1 and g1 were 0.49 and 0.465,respectively.

This results (for each piece type, move number, set and exper-tise level) in an 8∗8 matrix which encodes the normalized averageoccupation of the corresponding piece along the board. Then,for each piece and move number, we compared the distribu-tions along the board for high and low rating, with indepen-dent two-sample t-tests for each entry of the matrix. Positivet-values indicate that the occupation probability in a given entryis more likely for high than for low rated players. p-values werecorrected with a strict Bonferroni criterion for multiple compar-isons, considering a difference as significant only if p < 0.001/64.We reported results from knights, rooks and queen in the fullmatrix.

Data is represented as mean ± SD (n = 35) for Figures 1C,2A,B. Asterisks indicate significant differences at p < 0.001(Bonferroni corrected). The same color code is maintained alongthe whole work: red indicates higher probabilities for strongplayers and blue for weaker players.

RESULTSHYPOTHESIS 1Low rated players make moves which are more proximal to theirown last move and to the opponent precedent move.

D1 considers the difference between the initial location of apiece movement and the final location of the previous movementby the opponent. D2 corresponds to the difference between thefinal locations of two successive moves from the same player (seeMaterials and Methods and Figure 1A for a full description ofhow D1 and D2 are calculated).

For both expertise groups, we found a similar distribution ofdistances probabilities: all players tended to make their move-ments in squares close to the final location of their opponentlast movement (D1) and to their own last movement (D2)(Figure 1B). However, and in accordance with our first hypothe-sis, lower rated players tend to make more movements in boardlocations which are proximal to the opponent’s last movement(D1) and their own antecedent movement (D2). Instead, strongplayers showed higher densities of successive movements which






FIGURE 2 | Object-based mechanisms depend on expertise level. (A)

Weak players repeat the same piece on consecutive movements morefrequently than strong players. Probabilities to repeat the same piece ontwo consecutive moves depends on the remaining number of pieces onthe board. We plotted this probability vs. the number of pieces for bothexpertise levels. Independent two-sample t-tests (high vs. low expertiseprobabilities) for each number of pieces left (ranging from 1 to 16) evidencea higher probability to move the same piece on consecutive movements forlow rated players, almost independently of the number of remaining pieces(∗p < 0.001 Bonferroni corrected for multiple comparisons). Dashed blackline shows the random threshold. (B) Low-rated players reduce the numberof pieces more rapidly than high-rated players. The number of remainingpieces over the board, which change throughout the game (starting in 16pieces for each player), is significantly higher for high rated players alongthe whole game (moves 3–40, independent two-sample t-tests on eachmove number, ∗p < 0.001 Bonferroni corrected for multiple comparisons),evidencing than low rated players exchange (or loss) their pieces morerapidly than stronger players.

were more scattered in space (Figure 1C) 1. To quantify thisobservation we aggregated the distribution of spatial differencesto a single scalar estimating the radial distance between the twosuccessive movements (Figure 1D).

1Note that D2 has a non-zero probability in the origin which corresponds totwo consecutive moves by one player which have the same end-location. Thiscan only happen when a player captures twice in the same location. Instead D1

has a strict zero probability in the origin since a player cannot make a movestarting in the square where the opponent has moved a piece.

Results of D1 indicated that movements starting within aradius of 3.5 squares from the position where the opponentmoved his piece were more likely in low rated players (p < 0.001Bonferroni corrected) while movements outside of this radiuswhere more likely for high rated players (p < 0.001 Bonferronicorrected). Similarly, results of D2 indicated that low rated play-ers were more likely (p < 0.001 Bonferroni corrected) to maketwo consecutive moves with end-locations within a radius of2.5 squares, while higher rated players made more likely movesbeyond this radius.

Both of these results are extremely reliable through distances(Figure 1D). This indicates that while the specific geometry ofthe board may be subtle and specific to chess (the patterns ofFigure 1C are complex) they organize on a synthetic rule bywhich low rated players tend to overplay more proximal and highrated players more distal moves.

Previous results were obtained using games with a short timebudget (180 s). To avoid any confound related with the use of veryshort games, we repeated the exact analysis for games with longertime budgets (300 and 900 s per player) and we found the sameresults (Figures S1, S2).

HYPOTHESIS 2Low rated players are more likely to move the same piece inconsecutive turns.

The previous results suggest that low rated players have a nar-rower (or more focal) spatial window of attention. Attention canalso be directed to objects (Richard et al., 2008). We examine thehypothesis that throughout a game, low rated players are morefocused in a specific piece than high rated players, who may driveattention to schemas assembling sets of pieces (pawn chains, sev-eral pieces converging in a square or a plan . . . ). To this aim,we simply measured the probability of repeating a piece in twoconsecutive moves. A repetition is counted only when the exactsame piece (not the same type of piece) is moved twice. If a playermoves a pawn and in the next turn moves another pawn, this isconsidered as a different piece movement.

The probability of moving the same piece twice depends on thenumber of pieces remaining of the board. Thus, we calculated therepetition probability as a function of the pieces remaining in theboard (Figure 2A). The repetition probability decreased with thenumber of pieces for both groups but remained above chance lev-els. This is expected since: (1) some pieces actually cannot moveand (2) players rarely consider all pieces in the board as candidatesto move. As we had hypothesized, lower rated players producedmore repetitions reflecting that attention (or their strategies orconception of plans) is more likely to be constrained to a singlepiece. To quantify this observation we first submitted the data toa Two-Way ANOVA test with number of pieces and expertise levelas independent factors and their interactions. Results showeda main effect of both factors (Expertise, p < 0.0001, F = 91.8,df = 1; Number of Pieces: p < 0.0001, F = 181.6, df = 15 andInteraction, p < 0.0001, F = 7.5, df = 15). We followed this testwith independent two sample t-tests (corrected with a strictBonferroni criterion for multiple comparisons) for each num-ber of pieces left, comparing the distributions for high and lowrated players. Each value of the distribution is obtained from one





of the 35 different sets of each expertise level. All comparisonsconsistently showed greater repetition probability for lower thanhigher rated player. This effect was significant when 6, 8–14, and16 pieces remaining on the board [t(34) < −5.2, p < 0.001].

As for the first hypothesis, we repeated the piece repetitionanalysis using games with longer time budgets (Figures S3A,B)and we replicated the results found for 180 s games.

HYPOTHESIS 3Since weak players focus more on individual pieces it is expectedthat it is effortful for them to work on boards with many pieces.We expect that weaker players will tend to simplify the position toavoid mental effort (Koechlin and Summerfield, 2007).

We compared the amount of pieces as a function of movenumber for both expertise levels. As expected the numberof pieces started in 16 (initial configuration) and smoothlydecreased to an average of about 7 pieces by move 40. Beyond thismain trend we observed that after the first five moves, the distri-butions of remaining pieces for high and low rated players bifur-cate. This analysis shows that, as we had hypothesized, weakerplayers exchange pieces more rapidly than stronger players. Toquantify this observation we first made a Two-Way ANOVA testwith number of pieces and expertise level as independent factorsand their interactions. Results showed a main effect of both fac-tors (Expertise, p < 0.0001, F = 4.5∗104, df = 1; Move number:p < 0.0001, F = 2.2∗105, df = 39 and Interaction, p < 0.0001,F = 296, df = 39). We followed this test with independent two-sample t-tests (corrected with a strict Bonferroni criterion formultiple comparisons) for each move number, comparing thedistributions for high and low rated players. All comparisons con-sistently showed significant greater number of pieces along thewhole game (move numbers 3–40) for high rated players [t(34) >

15 for move numbers 4–6, t(34) > 27 for move numbers 3 and7–40, p < 0.001]. Once again, we repeated the previous analysisfor 300 and 900 s games (Figures S3C,D) replicating the resultsfound for 180 s time budget.

The results described above show that weaker players tend toproduce consecutive moves in proximal board locations, moreoften moving the same piece and exchanging pieces more rapidlyto reduce the number of remaining pieces. These three principlesreflect consistent general findings which might reflect the effect ofexpertise on human actions in complex setups.

Beyond these three principles there are idiosyncratic aspects ofthe game of chess which relate to piece value, with the way themove on the board and where they are more effective, which dic-tate a specific strategy in the board. Expert play is expected tobe more in-line with certain strategic themes. Here we examinedthree core aspects of chess strategy: (1) Knights are more effectivewhen they are centralized, (2) Rooks play first along the 1st rowto find an optimal centralized column where they are effective,(3) The queen should not risk going for long travels early in thegame.

At different stages of the game (moves 5, 10, 20, 30, and 40)we calculated the average density of white pieces (N: knights,R: Rooks, Q: Queen) along the board. For each square we per-formed a two-sample t-test comparing the distribution of densi-ties for the high and low rated players and corrected for multiple

comparisons with a strict Bonferroni criterion. Results showedthat, as expected, knights were significantly more centralized(over the whole board) for higher rated players and were morelikely to be found in their initial square (b1 or g1) or advancedin the enemy camp for weaker players (Figure 3A). Rook move-ment revealed that by move 5 higher rated players are more likelyto have castled and by move 10 a higher probability of central-izing the rooks on the 1st row (Figure 3B). Rooks position inweaker players instead was much more likely to be in the ini-tial squares (a1 and h1). Also, as with knights, weaker playersare more likely to advance the rook in the enemy camp. Anotherconsistent finding is that stronger players place their rooks in thequeen-side (left side of the board for white). Instead weaker play-ers more rapidly attack on the king-side: for instance by move30, white rooks are more likely to be found in squares close tothe opponent-king location. This reflects a more positional playand a less direct tendency of directly going to mate the enemyking for high rated players. Finally, as expected, stronger play-ers tend to postpone queen development (by move 5 and 10 thequeen is more likely to be in the initial square d1) and through-out the game develop the queen on the first rows (Figure 3C). Weemphasize that these results do not convey information about theabsolute distribution of occupation of a piece. Instead they reflectdifferential distributions, i.e., indicating whether a given piecesis more likely to be occupied by stronger or weaker players. Toavoid confusion Figure S4 shows the average degree of occupancyof these pieces throughout the game for each expertise level. Asfor previous hypotheses, we replicated and found similar resultsfor games with longer time budgets (see Figure S5).

DISCUSSIONHere we showed differences in general and domain-specific pat-terns of actions depending on expertise level. We found thatweaker players play more locally, tend to focus sequences ofactions in the same piece and more rapidly exchange pieces toreduce the total number of pieces on the board. Our workinghypothesis is that these observations reflect a different focus ofattention to space and objects with expertise.

Our first working hypothesis was that the end location of amovement (the place where a piece is located) functions like aspatial cue in the board space. Our conjecture is that the ten-dency to continue playing close to that cue is a reflection of thepersistence of attention to this location.

Attention involves a sequence of operations: 1- attentionalshifting to the target location, 2-attentional engagement, 3-attentional disengagement (Posner and Petersen, 1990; Fox et al.,2002; Koster et al., 2006). The persistence of play in a givenlocation has three likely and related explanations: a first possi-bility is that novices have difficulties disengaging attention fromthe cued place in the last movement (Sheridan and Reingold,2013). Another possibility is that expert players have more effec-tive preattentive mechanisms to encode saliency mechanisms inperipheral locations of the board (Sigman and Gilbert, 2000;Intriligator and Cavanagh, 2001; Pylyshyn, 2007; Vinckier et al.,2007). Chess masters have an advantage for the recognition ofchess pieces (Saariluoma, 1995; Kiesel et al., 2009; Bilalic et al.,2010) and chess themes (Reingold et al., 2001a,b), which may






FIGURE 3 | Piece distribution over the board reveals domain-specific and

expertise-dependent strategies. Probabilities to find a type of piece(normalized by the number of remaining pieces of this type) in each square(8∗8 checkerboard) were compared for high and low expertise groups atdifferent game stages (moves 5, 10, 20, 30, and 40). The t-value resulting ofthe two-sample t-test (high vs. low expertise group), in each square is colorcoded for those significantly different comparisons with p < 0.001 Bonferronicorrected for multiple comparisons. Red positive values indicate significantlyhigher probabilities for strong players and blue negative values, significantly

higher probabilities for weaker players. (A) Knights, (B) Rooks, and (C) Queenoccupancy comparisons reveal that strong players centralize more theirknights and rooks, delaying the queen development, compared with lowerrated players delaying more the development of knights and rooks (but notthe development of the queen, which is almost always not good) and/oroccupying more advanced squares with all. It should be noted that the thisfigure represents the differential occupancy of each square, showing thatstrong or weak players locate each type of piece comparatively morefrequently than the other group.

serve as salient detector providing new cues which compete withthe previous spatial cue. A third possibility is that expert playerscan attend to themes or schemas (strings of pieces) and the focusof attention is spread over the whole attended object (Houtkampet al., 2003; Alvarez and Scholl, 2005; Richard et al., 2008). Inline with this idea, the size and form of the selection windowhas been proposed to be controlled by top-down mechanismsand dependent on the task difficulty (Belopolsky and Theeuwes,2010).

Our results based on the spontaneous distributions of actionsduring a game are consistent with a recent report by Sheridan andReingold analyzing the distribution of attention in the EinstellungEffect (Sheridan and Reingold, 2013). The authors present a prob-lem in which there is a move which almost indefectibly attractsattention (for instance a region on the board where there seemsto be mate, with the king exposed and many pieces attackingit). In this construction, the best move is away from this specificlocation of the board. Weaker players very often do not con-sider this optimal (but distant from a very salient location) moveand their gaze remains in the Einstellung region of the board.Instead, stronger players can disengage from this location whichallows them to find the distant and optimal move (Sheridan andReingold, 2013).

Weaker players are less likely to organize the representationof the board in large chunks (Chase and Simon, 1973; Gobetand Simon, 1996) and thus, comparing variations when manypieces remain on the board requires more attentional shiftsand executive function. Koechlin and colleagues have coined theidea of a “lazy” executive system which is triggered only whenstrictly needed (Koechlin and Summerfield, 2007). Combining

these premises, we reasoned that all players will seek to min-imize effort. To achieve this, weaker players will prefer posi-tions with less number of pieces on the board. As expected bythis prediction we show that weak players tend to “simplify”the problem by more rapidly removing pieces of the boardconsistently.

We initially chose using 180 s games to test our hypothesesbased on: 1- evidences showing that rapid processes (related withpattern recognition) rather than slow processes (fundamentally,search mechanisms) are responsible of chess expertise (Burns,2004; Sigman et al., 2010), 2- the time used for each movementis close-related with the common psychological experiments and3- the FICS database for 180 s games is larger than those forlonger time budgets. However, we also replicated all our resultsfor longer time budgets (300 and 900 s) indicating the robust-ness of the conclusions and absence of 180 s time budget-relatedartifacts.

The effect size of all our results is small but consistent withour working hypotheses and each result was replicated in threedifferent time budgets, favoring the consistency and reliability ofour results.

Our study focused on chess but there is no reason to thinkthat the main of the conclusions derived here (with the excep-tion of specific strategic patterns shown in Figure 3) are specificto chess and hence are likely be generalized to the effect of exper-tise on other domains of human action and decision making (i.e.,novice car drivers focusing his/her attention only in the front roadbut not in the car mirrors or novice sport players deciding theiractions based only in the ball location because they are not able tosimultaneously attend to their partners and opponents locations).





ACKNOWLEDGMENTSThis research was supported by the Consejo Nacional deInvestigaciones Científicas y Técnicas (CONICET). MarianoSigman is sponsored by the James McDonnell Foundation.


Figure S1 | Spatial effects are also observed in longer games: 300 s time

budget. (A) Probability distributions of movement distances. As for 180 s

games (Figure 1B), probabilities to make a movement close to the

previous one are higher at short distances, for both expertise levels. (B)

Weak players made their movements closer to the previous one. The

same analysis explained for 180 s games was made for 300 s games,

obtaining similar results (see Figure 1C). Probability distributions for both

High and Low rated players were contrasted on each entry of the 15∗15

distance square independently. t-value of each independent two-sample

t-test (with p-value < 0.001, Bonferroni corrected for multiple

comparisons) is color-coded. Positive (red) t-values indicate significantly

higher probabilities for high rated players and negative (blue) values, for

weaker players. (C) Radial or Euclidean distances. Again, as in Figure 1D

for 180 s games, distances were one-dimension collapsed and the

difference between probabilities of making movements corresponding to

a distance square [P(High) – P(Low)] was plotted vs. each radial distance.

High and low rated groups distributions were independently compared in

each radial distance [two-sample t-test on each variable (D1 and D2)]. Red

asterisks (t > 5.2 for D1 and t > 4.9 for D2) indicates distances were

P(High) is significantly higher than P(Low); blue asterisks (t < −7.3 for D1

and t < −7.8 for D2), distances were P(Low) > P(High); in both p < 0.001,

Bonferroni corrected for multiple comparisons.

Figure S2 | Spatial effects are also observed in longer games: 900 s time

budget. (A) Probability distributions of movement distances. As for 180 s

(Figure 1B) and 300 s (Figure S1A) games, probabilities to make a

movement close to the previous one are higher at short distances (for

both expertise levels). (B) Weak players made their movements closer to

the previous one. The same analysis explained for 180 s and 300 s games

was made for 900 s games, obtaining similar results (see Figure 1C).

Probability distributions for both high and low rated players were

contrasted on each entry of the 15∗15 distance square independently.

t-value of each independent two-sample t-test (with p-value < 0.001,

Bonferroni corrected for multiple comparisons) is color-coded. Positive

(red) t-values indicate significantly higher probabilities for high rated

players and negative (blue) values, for weaker players. (C) Radial or

Euclidean distances. Again, as for 180 s games (Figure 1D) and 300 s

games (Figure S1C), distances were one-dimension collapsed and the

difference between probabilities of making movements corresponding to

a distance square [P(High) – P(Low)] was plotted vs. each radial distance.

High and low rated groups distributions were independently compared in

each radial distance [two-sample t-test on each variable (D1 and D2)]. Red

asterisks (t > 6.9 for D1 and t > 4.9 for D2) indicates distances were

P(High) is significantly higher than P(Low); blue asterisks (t < −5.3 for D1

and t < −5 for D2), distances were P(Low) > P(High); in both p < 0.001,

Bonferroni corrected for multiple comparisons.

Figure S3 | For longer games (300 and 900 s time budget), object-based

mechanisms also depend on expertise level. (A,B) Weak players repeat

the same piece on consecutive movements more frequently than strong

players. As for 180 s games (Figure 2A), we plotted the probability to

repeat the same piece on successive movements vs. the number of

pieces for both expertise levels for 300 and 900 s games. To quantify

these observations we first submitted the data to a Two-Way ANOVA test

with number of pieces and expertise level as independent factors and

their interactions. Results for 300 s games showed a main effect of both

factors (Expertise, p < 0.0001, F = 41.9, df = 1; Number of Pieces:

p < 0.0001, F = 61.6, df = 15 and Interaction, p < 0.0001, F = 10.1,

df = 15). Results for 900 s games showed a main effect of both factors

(Expertise, p < 0.0001, F = 197.8, df = 1; Number of Pieces: p < 0.0001,

F = 89.7, df = 15 and Interaction, p < 0.0001, F = 31.4, df = 15). We

followed these tests with independent two sample t-tests (corrected with

a strict Bonferroni criterion for multiple comparisons) for each number of

pieces left, comparing the distributions for high and low rated players.

Each value of the distribution is obtained from one of the 35 different sets

of each expertise level. All comparisons consistently showed greater

repetition probability for lower than higher rated player. For 300 s games,

this effect was significant for 7–14 pieces remaining on the board

[t(34) < −4.9, p < 0.001]. For 900 s games, this effect was significant for

8–16 pieces remaining on the board [t(34) < −6.4, p < 0.001]. Dashed

black line shows the random threshold. (C,D) Low-rated players reduce

the number of pieces more rapidly than high-rated players. As it was

previously showed for 180 s games, the number of remaining pieces over

the board is significantly higher for high rated players almost for the whole

game for both 300 and 900 s games. First, we made a two-way ANOVA

test with number of pieces and expertise level as independent factors and

their interactions showed a main effect of both factors. ANOVA results for

300 s games: Expertise, p < 0.0001, F = 1.3∗104, df = 1; Move number:

p < 0.0001, F = 3.2∗104, df = 39 and Interaction, p < 0.0001, F = 175,

df = 39. ANOVA results for 900 s games: Expertise, p < 0.0001,

F = 0.98∗104, df = 1; Move number: p < 0.0001, F = 2.2∗104, df = 39

and Interaction, p < 0.0001, F = 79, df = 39. We followed these tests

with independent two-sample t-tests (corrected with a strict Bonferroni

criterion for multiple comparisons) for each move number, comparing the

distributions for high and low rated players. All comparisons consistently

showed significant greater number of pieces almost along the whole

game [in 300 s games, move numbers 3, and 8–40, t(34) > 5, p < 0.001; in

900 s games, move numbers 3 and 7–40, t(34) > 6, p < 0.001] for high

rated players, evidencing than low rated players exchange (or loss) their

pieces more rapidly than stronger players in 300 and 900 s games.

Figure S4 | Occupation distribution of pieces throughout the game for

both expertise levels. Probabilities to find a (A) Knight, a (B) Rook, or the

(C) Queen at each square of the chess board is represented for each

rating group at different game stages (move numbers 5, 10, 20, 30, and

40). It should be noted that both groups occupy almost the same squares

along the board (there are not “exclusive” squares), but some places are

comparatively more occupied by weak or strong players (see Figure 3).

Figure S5 | Piece distribution over the board also reveals domain-specific

and expertise-dependent strategies in longer games (300 and 900 s time

budget). Probabilities to find a type of piece (normalized by the number of

remaining pieces of this type) in each square (8∗8 checkerboard) were

compared for high and low expertise groups at different game stages

(moves 5, 10, 20, 30, and 40). The t-value resulting of the two-sample

t-test (high vs. low expertise group), in each square is color coded for

those significantly different comparisons with p < 0.001 Bonferroni

corrected for multiple comparisons. Red positive values indicate








significantly higher probabilities for strong players and blue negative

values, significantly higher probabilities for weaker players. (A,D) Knights,

(B,E) Rooks, and (C,F) Queen occupancy comparisons reveal that strong

players centralize more their knights and rooks, delaying the queen

development, compared with lower rated players delaying more the

development of knights and rooks (but not the development of the queen,

which is almost always not good) and/or occupying more advanced

squares with all. It should be noted that the this figure represents the

differential occupancy of each square, showing that strong or weak

players locate each type of piece comparatively more frequently than the

other group. (A–C) correspond to 300 s games. (D–F) correspond to 900 s

games.

REFERENCESAlvarez, G. A., and Scholl, B. J. (2005). How does attention select and track spatially

extended objects? New effects of attentional concentration and amplification.J. Exp. Psychol. Gen. 134, 461–476. doi: 10.1037/0096-3445.134.4.46

Belopolsky, A. V., and Theeuwes, J. (2010). No capture outside the attentionalwindow. Vision Res. 50, 2543–2550. doi: 10.1016/j.visres.2010.08.023

Bilalic, M., Langner, R., Erb, M., and Grodd, W. (2010). Mechanisms and neu-ral basis of object and pattern recognition: a study with chess experts. J. Exp.Psychol. Gen. 139, 728–742. doi: 10.1037/a0020756


Burns, B. D. (2004). The effects of speed on skilled chess performance. Psychol. Sci.15, 442–447. doi: 10.1111/j.0956-7976.2004.00699.x

Campitelli, G., and Gobet, F. (2008). The role of practice in chess: a longitudinalstudy. Learn. Individ. Differ. 18, 446–458. doi: 10.1016/j.lindif.2007.11.006

Cepeda, N. J., Cave, K. R., Bichot, N. P., and Kim, M. S. (1998). Spatial selec-tion via feature-driven inhibition of distractor locations. Percept. Psychophys.60, 727–746. doi: 10.3758/BF03206059

Chase, W. G., and Simon, H. A. (1973). Perception in chess. Cogn. Psychol. 4, 55–81.Chun, M. M., and Jiang, Y. (1998). Contextual cueing: implicit learning and mem-

ory of visual context guides spatial attention. Cogn. Psychol. 36, 28–71. doi:10.1006/cogp.1998.0681

Connors, M. H., Burns, B. D., and Campitelli, G. (2011). Expertise in complex deci-sion making: the role of search in chess 70 years after de Groot. Cogn. Sci. 35,1567–1579. doi: 10.1111/j.1551-6709.2011.01196.x

de Groot, A. D. (1978). Thought and Choice in Chess (2nd. Edn.). New York, NY:Mounton De Gruyter.

Duncan, J. (1984). Selective attention and the organization of visual information.J. Exp. Psychol. Gen. 113, 501–517. doi: 10.1037/0096-3445.113.4.501

Elo, A. E. (1978). The Rating of the Chess Players, Past and Present. New York, NY:Arco.

Fox, E., Russo, R., and Dutton, K. (2002). Attentional bias for threat: evidence fordelayed disengagement from emotional faces. Cogn. Emot. 16, 355–379. doi:10.1080/02699930143000527

Gilbert, C. D., and Sigman, M. (2007). Brain states: top-down influences in sensoryprocessing. Neuron 54, 677–696. doi: 10.1016/j.neuron.2007.05.019


Houtkamp, R., Spekreijse, H., and Roelfsema, P. R. (2003). A gradual spread ofattention during mental curve tracing. Percept. Psychophys. 65, 1136–1144. doi:10.3758/BF03194840

Intriligator, J., and Cavanagh, P. (2001). The spatial resolution of visual attention.Cogn. Psychol. 43, 171–216. doi: 10.1006/cogp.2001.0755

Kiesel, A., Kunde, W., Pohl, C., Berner, M. P., and Hoffmann, J. (2009). Playingchess unconsciously. J. Exp. Psychol. Learn. Mem. Cogn. 35, 292–298. doi:10.1037/a0014499

Koechlin, E., and Summerfield, C. (2007). An information theoretical approachto prefrontal executive function. Trends Cogn. Sci. 11, 229–235. doi:10.1016/j.tics.2007.04.005

Koster, E. H., Crombez, G., Verschuere, B., Van Damme, S., and Wiersema, J. R.(2006). Components of attentional bias to threat in high trait anxiety: facilitatedengagement, impaired disengagement, and attentional avoidance. Behav. Res.Ther. 44, 1757–1771. doi: 10.1016/j.brat.2005.12.011

Maunsell, J. H., and Treue, S. (2006). Feature-based attention in visual cortex.Trends Neurosci. 29, 317–322. doi: 10.1016/j.tins.2006.04.001

New, J., Cosmides, L., and Tooby, J. (2007). Category-specific attention for ani-mals reflects ancestral priorities, not expertise. Proc. Natl. Acad. Sci. U.S.A. 104,16598–16603. doi: 10.1073/pnas.0703913104

Posner, M. I., and Petersen, S. E. (1990). The attention system of the humanbrain. Annu. Rev. Neurosci. 13, 25–42. doi: 10.1146/annurev.ne.13.030190.000325

Pylyshyn, Z. W. (2007). Things and Places: How the Mind Connect With the World.London: MIT Press.

Reingold, E. M., Charness, N., Pomplun, M., and Stampe, D. M. (2001a). Visualspan in expert chess players: evidence from eye movements. Psychol. Sci. 12,48–55. doi: 10.1111/1467-9280.00309

Reingold, E. M., Charness, N., Schultetus, R. S., and Stampe, D. M. (2001b).Perceptual automaticity in expert chess players: parallel encoding of chessrelations. Psychon. Bull. Rev. 8, 504–510. doi: 10.3758/BF03196185

Richard, A. M., Lee, H., and Vecera, S. P. (2008). Attentional spreading in object-based attention. J. Exp. Psychol. Hum. Percept. Perform. 34, 842–853. doi:10.1037/0096-1523.34.4.842

Saariluoma, P. (1995). Chess Player’s Thinking: A Cognitive Psychological Approach.London: Routledge.

Schultetus, R. S., and Charness, N. (1999). Recall or evaluation of chess positionsrevisited: the relationship between memory and evaluation in chess skill. Am. J.Psychol. 112, 555–569. doi: 10.2307/1423650

Sheridan, H., and Reingold, E. M. (2013). The mechanisms and boundary con-ditions of the Einstellung effect in chess: evidence from eye movements. PLoSONE 8:e75796. doi: 10.1371/journal.pone.0075796

Sigman, M., Etchemendy, P., Slezak, D. F., and Cecchi, G. A. (2010). Response timedistributions in rapid chess: a large-scale decision making experiment. Front.Neurosci. 4:60. doi: 10.3389/fnins.2010.00060

Sigman, M., and Gilbert, C. D. (2000). Learning to find a shape. Nat. Neurosci. 3,264–269. doi: 10.1038/72979

Tatler, B. W., Wade, N. J., Kwan, H., Findlay, J. M., and Velichkovsky, B. M. (2010).Yarbus, eye movements, and vision. Iperception 1, 7–27. doi: 10.1068/i0382

Vinckier, F., Dehaene, S., Jobert, A., Dubus, J. P., Sigman, M., and Cohen, L.(2007). Hierarchical coding of letter strings in the ventral stream: dissecting theinner organization of the visual word-form system. Neuron 55, 143–156. doi:10.1016/j.neuron.2007.05.031

Yarbus, A. (1967). Eye Movements and Vision. New York, NY: Plenum Press.


Received: 11 December 2013; accepted: 15 January 2014; published online: 04 February2014.Citation: Leone MJ, Fernandez Slezak D, Cecchi GA and Sigman M (2014) Thegeometry of expertise. Front. Psychol. 5:47. doi: 10.3389/fpsyg.2014.00047This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Leone, Fernandez Slezak, Cecchi and Sigman. This is an open-access article distributed under the terms of the Creative Commons Attribution License(CC BY). The use, distribution or reproduction in other forums is permitted, providedthe original author(s) or licensor are credited and that the original publication in thisjournal is cited, in accordance with accepted academic practice. No use, distribution orreproduction is permitted which does not comply with these terms.












GENERAL COMMENTARYpublished: 03 April 2014

doi: 10.3389/fpsyg.2014.00270

Expertise and the representation of spaceMichael H. Connors1,2* and Guillermo Campitelli 3

1 ARC Centre of Excellence in Cognition and its Disorders, Department of Cognitive Science, Macquarie University, Sydney, NSW, Australia2 Dementia Collaborative Research Centre, School of Psychiatry, University of New South Wales, Sydney, NSW, Australia3 School of Psychology and Social Science, Edith Cowan University, Perth, WA, Australia*Correspondence: [email protected]

Edited by:

Merim Bilalic, Alpen Adria University Klagenfurt, Austria

Reviewed by:

Maria Juliana Leone, University of Buenos Aires, Argentina

Keywords: big data, chess, chess expertise, chunks, expertise, decision making, methodology

A commentary on

The geometry of expertiseby Leone, M. J., Slezak, D. F., Cecchi, G.A., and Sigman, M. (2014). Front. Psychol.5:47. doi: 10.3389/fpsyg.2014.00047

Across many domains, experts makedecisions based on the spatial relation-ships of objects within an environment.Firefighters, for example, need to evalu-ate the fire in front of them, radiologiststhe medical scan, and chess players theposition on the chess board before mak-ing a decision. In order to be effective,experts need to assess these spatial rela-tionships quickly and despite possibleuncertainty. This ability is facilitated bycognitive processes. According to the so-called chunking and template theories,this expertise is made possible by pat-tern recognition (Gobet and Simon, 1996;Gobet, 1997). Based on their experienceand practice, experts store in their long-term memory meaningful configurationsof elements (i.e., patterns—chunks ortemplates) in their domain of expertise.These patterns are associated with typicaldecisions and strategies. Thus, when anexpert faces a situation in their domainof expertise, they are able to rapidly rec-ognize patterns, which, in turn, promptthem to consider decisions or strategiesthat have been effective in previous situa-tions. This allows experts to make superiordecisions to novices who are limited to areal-time search through a large numberof possible options (Feltovich et al., 2006;Gobet and Charness, 2006; Connors et al.,2011).

Leone et al.’s (2014) elegant study addsto existing evidence for this account. Inparticular, Leone et al. demonstrate thatexperts’ representation of space in theirdomain of expertise is qualitatively differ-ent to that of novices. Leone et al. focusedon chess. Chess has the advantage of beinga constrained task environment with rel-atively high ecological validity to otherdomains. Chess also has the advantage of aprecise measure of expertise in the form ofa numerical rating that is assigned to eachplayer on the basis of their performanceagainst their opponents. For these rea-sons, a large amount of expertise researchhas examined chess (Gobet and Charness,2006). For Leone et al.’s study, chess hasthe additional advantage of having largeamounts of data available: Leone et al.examined well over 175,000 games drawnfrom an internet chess server.

Given experts’ pattern recognition abil-ity, Leone et al. predicted that expertsshould make moves that were best for theposition at hand and so not be constrainedby the mere physical distance betweenobjects. In contrast, novices would bemore likely to make moves based on spatialproximity to previous moves due to theirrelative lack of pattern recognition andtheir need to conserve cognitive resources.Leone et al. made four specific hypotheses.First, the authors predicted that novicesshould make moves that are closer to theiror their opponent’s previous move. To testthis, they developed a novel methodol-ogy in which they calculated the distancebetween a player’s move and their previousmove, and the distance between a player’smove and their opponent’s previous move.

Second, the authors predicted that noviceplayers would be more likely to movethe same piece multiple times. To testthis, they examined the frequency withwhich players moved the same piece morethan once on consecutive turns. Third,the authors predicted that novices wouldbe more likely to simplify the positionby exchanging pieces (capturing an oppo-nent’s piece when the opponent can recap-ture their own piece), thereby reducingthe cognitive load of dealing with largenumbers of pieces. To test this, they exam-ined the rate at which pieces were removedfrom the board according to the num-ber of moves in the game. Finally, theauthors predicted that novices would beless likely to keep with general strategicprinciples (e.g., knights are more effec-tive when centralized) than experts. Totest this, they examined the frequencywith which pieces were placed in suitableparts of the board (e.g., knights in centralregions).

Leone et al. found evidence for allfour hypotheses. First, novices were morelikely to move pieces that were closer totheir or their opponent’s previous movethan experts. Second, novices were morelikely to move the same piece on consec-utive turns than experts. Third, noviceswere more likely to simplify positionsthan experts. Finally, novices were lesslikely to keep to general strategic princi-ples than experts. Although the relativesize of the effects were small, Leone et al.demonstrated the robustness of the effectsacross very large datasets and across gamesof different durations. Similar differencesbetween experts and novices were evident

www.frontiersin.org April 2014 | Volume 5 | Article 270 | 135












Connors and Campitelli Expertise and the representation of space

regardless of whether players had a total of3, 5, or 15 min each.

Although the authors did not directlyassess the quality of the players’ decisions,their work provides evidence that expertsbase their decision on what is most mean-ingful in a position, rather than beinglimited by spatial proximity. As Leoneet al. note, these findings are consis-tent with chunking and template mod-els of expertise as the differences betweenexperts and novices exist at very shorttime limits, when searching through var-ious options is not possible. The findingsare also likely to generalize to a wide rangeof other domains of expertise. In addi-tion to revealing these effects, the work isvery important because it develops novelmethodologies for assessing spatial rela-tionships in chess positions and acrosslarge datasets. Indeed, Leone et al.’s useof big data to test specific hypothesesabout cognition is particularly innovative.In future, these methods might provide

further insight into how experts interpretand represent space.

REFERENCESConnors, M. H., Burns, B. D., and Campitelli, G.

(2011). Expertise in complex decision making: therole of search in chess 70 years after de Groot.Cogn. Sci. 35, 1567–1579. doi: 10.1111/j.1551-6709.2011.01196.x

Feltovich, P. J., Prietula, M. J., and Ericsson, K.A. (2006). “Studies of expertise from psy-chological perspectives,” in The CambridgeHandbook of Expertise and Expert Performance,eds K. A. Ericsson, N. Charness, P. J.Feltovich, and R. R. Hoffman (Cambridge:Cambridge University Press), 41–68. doi:10.1017/CBO9780511816796.004

Gobet, F. (1997). A pattern-recognition theory ofsearch in expert problem solving. Think. Reason.3, 291–313. doi: 10.1080/135467897394301

Gobet, F., and Charness, N. (2006). “Expertise inchess,” in The Cambridge Handbook of Expertiseand Expert Performance, eds K. A. Ericsson, N.Charness, P. J. Feltovich, and R. R. Hoffman(Cambridge: Cambridge University Press),523–538. doi: 10.1017/CBO9780511816796.030

Gobet, F., and Simon, H. A. (1996). Templatesin chess memory: a mechanism for recalling

several boards. Cogn. Psychol. 31, 1–40. doi:10.1006/cogp.1996.0011

Leone, M. J., Slezak, D. F., Cecchi, G. A., andSigman, M. (2014). The geometry of expertise.Front. Psychol. 5:47. doi: 10.3389/fpsyg.2014.00047


Received: 26 February 2014; accepted: 12 March 2014;published online: 03 April 2014.Citation: Connors MH and Campitelli G (2014)Expertise and the representation of space. Front. Psychol.5:270. doi: 10.3389/fpsyg.2014.00270This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Connors and Campitelli. This isan open-access article distributed under the terms ofthe Creative Commons Attribution License (CC BY).The use, distribution or reproduction in other forumsis permitted, provided the original author(s) or licen-sor are credited and that the original publicationin this journal is cited, in accordance with acceptedacademic practice. No use, distribution or repro-duction is permitted which does not comply withthese terms.

Frontiers in Psychology | Cognition April 2014 | Volume 5 | Article 270 | 136













doi: 10.3389/fpsyg.2014.00923

Playing off the curve - testing quantitative predictions ofskill acquisition theories in development of chessperformanceRobert Gaschler1*, Johanna Progscha2, Kieran Smallbone3, Nilam Ram4 and Merim Bilalic 5

1 Universität Koblenz-Landau, Landau, Germany and Interdisciplinary Research Laboratory Image, Knowledge, Gestaltung at Humboldt-Universität, Berlin, Germany2 Department of Psychology, Humboldt-Universität, Berlin, Berlin, Germany3 School of Computer Science, University of Manchester, Manchester, UK4 College of Health and Human Development, Human Development and Family Studies, Pennsylvania State University, University Park, PA, USA5 Alpen-Adria-Universität Klagenfurt, Institut für Psychologie, Abteilung für Allgemeine Psychologie und Kognitionsforschung, Klagenfurt, Austria

Edited by:

Michael H. Connors, MacquarieUniversity, Australia

Reviewed by:

Michael H. Connors, MacquarieUniversity, AustraliaEyal M. Reingold, University ofToronto, Canada

*Correspondence:

Robert Gaschler, Department ofPsychology, UniversitätKoblenz-Landau, Fortstraße 7,D-76829 Landau, Germanye-mail: [email protected]

Learning curves have been proposed as an adequate description of learning processes,no matter whether the processes manifest within minutes or across years. Differentmechanisms underlying skill acquisition can lead to differences in the shape of learningcurves. In the current study, we analyze the tournament performance data of 1383 chessplayers who begin competing at young age and play tournaments for at least 10 years.We analyze the performance development with the goal to test the adequacy of learningcurves, and the skill acquisition theories they are based on, for describing and predictingexpertise acquisition. On the one hand, we show that the skill acquisition theories implyinga negative exponential learning curve do a better job in both describing early performancegains and predicting later trajectories of chess performance than those theories implying apower function learning curve. On the other hand, the learning curves of a large proportionof players show systematic qualitative deviations from the predictions of either type ofskill acquisition theory. While skill acquisition theories predict larger performance gains inearly years and smaller gains in later years, a substantial number of players begin to showsubstantial improvements with a delay of several years (and no improvement in the firstyears), deviations not fully accounted for by quantity of practice. The current work addsto the debate on how learning processes on a small time scale combine to large-scalechanges.

Keywords: learning curves, skill acquisition, expertise, chess, development

INTRODUCTIONAnderson (2002) drew attention to the problem of time scalesin psychology with the programmatic article Spanning SevenOrders of Magnitude. On the one hand, acquisition of expertiseis known to takes years (e.g., Ericsson et al., 1993). On the otherhand, expertise research has a strong basis in cognitive psychologyparadigms wherein a large repertoire of laboratory tasks are usedto understand and chart changes in potential subcomponents ofexpertise acquisition over minutes or hours. This includes com-ponent skills such as verifying and storing chess patterns (Gobetand Simon, 1996a,b,c, 1998; Campitelli et al., 2005, 2007; Bilalicet al., 2009a), learning to discard irrelevant perceptual featuresfrom processing (e.g., Gaschler and Frensch, 2007; Reingold andSheridan, 2011) or overcoming dysfunctional bindings of knowl-edge structures (e.g., Bilalic et al., 2008a,b). Anderson suggestedthat while meaningful educational outcomes take at least tens ofhours to achieve, those outcomes can be traced back to opera-tions of attention and learning episodes at the millisecond level.He went beyond offering the perspective that expertise acquisitionshould in principle be reducible to small scale learning episodes.Rather, Anderson suggested that the problem of linking domains

of (a) laboratory cognitive psychology/neurocognitive researchand (b) educational/developmental science should be tractable,because small scale learning episodes would sum up to largescale developmental/educational changes of the same functionalform. Increases in overall performance as well as increases in effi-ciency of components (e.g., keystrokes, eye movements and factretrieval) over time are well described by the power function (seealso Lee and Anderson, 2001). Power functions of improvementsin simple components add up to a power-function improve-ment at the large scale. Scalability across time-scales would offerstraightforward linking of change taking place within minutes tochange taking place over years.

The power function (as well as the negative exponential func-tion, see Table 1 and Figure 1) describes negatively acceleratedchange of performance with practice. Early in practice, the abso-lute improvement in performance per unit of time invested islarge. Later on, the improvement per unit of time diminishes.Apart from improvements in hour-long laboratory learning tasks,the power function has been used to describe motor skills inindividuals differing in amount of practice on the scale of years(e.g., up to 7 years of cigar-rolling in Crossman, 1959, see Newell















Gaschler et al. Playing off the curve

et al., 2001 for an overview). Description of practice gains withthe power function are widespread in the literature (Newell andRosenbloom, 1981; Kramer et al., 1991; Lee and Anderson, 2001;Anderson, 2002) and consistent with prominent models of skilledperformance such as ACT-R (Anderson, 1982) or the instancemodel of automatization (Logan, 1988, 1992).

However, the reason for the dominance of the power func-tion in describing the functional form of describing practice hasbeen debated in the literature on skill acquisition. Heathcote et al.(2000) see also Haider and Frensch (2002) argued that the analysisof averaged data favors the power over the exponential functionas a statistical artifact. They suggested computation of power andexponential curves with non-aggregated data, separately for eachparticipant. They found an advantage of the negative exponentialfunction over the power function in 33 of 40 different re-analyzeddata sets with an average improvement in fit of 17%. Note thatsuccess of a mathematical model in fitting data better than a com-petitor model might not mean that it provides a more concisedescription. Potentially, one mathematical model is more flexiblethan the other, and better able to accommodate systematic as well

Table 1 | Formulas of negative exponential and power function and

the example parameters used for Figure 1.

Formula Negative Power

exponential function

function

Rating = A − B * Rating = A − B*

exp(−C * T) T** − C

Parameterdescriptionsand numericalexamples fromFigure 1

A = asymptote 1830 2200B = differencebetweenasymptote andinitialperformance

790 1000

C = curvature 0.18 0.3

T = time Year 1–20 Year 1–20

as chance features in the data. Thus, further credence is lent to amodel by accurate prediction rather than fitting (i.e., without anyfurther parameter adjustments; cf. Roberts and Pashler, 2000; Pittet al., 2002; Wagenmakers, 2003; Marewski and Olsson, 2009).

It is worthwhile considering the exact shape of the learningcurve to predict future performance. Furthermore, the differ-ences between exponential and power function are linked toassumptions in theories of skill acquisition (see below). Figure 1represents schematic examples of learning curves and deriva-tives. The left panel depicts a power function and an exponentialfunction that start at the same level in the first year of chess tour-nament participation and approach similar levels in year 20 oftournament participation. The power function shows especiallystrong performance gains in the first years. For instance, the gainin rating points (e.g., Elo, 1978) in year one is about double thesize of the gain in year two. Year two still yields considerably moreperformance gain as compared to year three, and so on and soforth. Absolute gain per year is depicted in the right panel. It isdecreasing for both, the power and the exponential function. Thequalitative difference between the two types of learning curvesbecomes most obvious when considering the relative learning rate(RLR). This rate is decreasing for the power function, but remainsconstant for the exponential function. In our example, the expo-nential function has a relative learning rate of about 20%. In eachyear, the players gain about 20% of the ELO points they havenot gained yet. If someone starts with 1000 and will end up with1500 points (see Method for an explanation of the scale used inchess), this would mean a gain of 100 points for the first year and80 points in the second year (20% of 1500 − (1000 + 100) = 80points).

One qualitative aspect of learning curves is that they rep-resent the diminishing absolute payoff of practice-investment.Exponential practice functions can be derived from a narrowset of assumptions. As Heathcote et al. (2000) explained oneneeds only to assume that learning is proportional to the timetaken to execute the component in case of a continuous mech-anism. First, a component that takes longer to execute presentsmore opportunity for learning. Second, as learning proceeds,the time to execute the component decreases. Therefore, the

800

1000

1200

1400

1600

1800

2000

1 2 3 4 5 6 7 8 9 1011121314151617181920

Ra�ng

Years played

Exponen�al

Power

0

20

40

60

80

100

120

140

160

180

200

. 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Gainperyear

Years played

Exp. abs. gain per year

Power abs. gain per year

Exp. % rela�ve gain per year

Power % rela�ve gain per year

A B

FIGURE 1 | (A) Schematic plot of the power function (blue lines) and negative exponential function (red lines) over 20 time points. (B) Shows the absolutedifferences in performance from one time point to the next (filled symbols) and the relative learning rate (empty symbols).






absolute learning rate decreases, resulting in exponential learning.Similarly, for discrete mechanisms, such as chunking, exponentiallearning can be explained by a reduction in learning opportu-nity. As responses are produced by larger and larger chunks,fewer opportunities for further composition are available. Time-demanding control is no longer necessary for small steps butonly for scheduling sets consisting of fixed series of small pat-terns. Naturally, the opportunities for compilation of small singleknowledge units into larger ones reduce, as more and morepatterns are already chunked.

Additional theoretical assumptions are needed to accommo-date a decreasing RLR. For instance, Newell and Rosenbloom(1981) see also Anderson (2002) assumed that chunks areacquired hierarchically and that every time a larger chunk is prac-ticed, this entails practice of its smaller components. Thus, bypracticing a knowledge unit consisting of sub-units, the sub-units and the overall pattern are fine-tuned and strengthened.Furthermore, at least in combinatorial environments, acquisitionproceeds ordered by chunk span. No larger span chunk is acquireduntil all chunks of smaller span have been acquired.

The above research suggests that one or the other simplelearning function might be adequate to describe improvementsover long time intervals (cf. Howard, 2014). Functions knownfrom work on short-term skill acquisition should be relevantto describe long-term expertise acquisition. We take chess as anexample to explore this perspective. First, longitudinal data span-ning years of practice are available. Second, theories on expertisein chess can be taken to suggest that scalability between small scalelearning and large scale expertise acquisition should be especiallylikely to hold in this domain. Expertise development in chessmight predominantly be based on cumulatively storing more andmore patterns of chess positions (Chase and Simon, 1973; Gobetand Simon, 1996b). Spatial (Waters et al., 2002; Connors andCampitelli, 2014; Leone et al., 2014) and perceptual capabilitiesare deemed crucial (Charness et al., 2001; Reingold et al., 2001;Bilalic et al., 2008a,b; Kiesel et al., 2009; Bilalic et al., 2010; Bilalicand McLeod, 2014). This suggests that attentional and learn-ing episodes taking place at the time scale of milliseconds mighttogether lead to expertise acquisition. This in turn would make itlikely that expertise acquisition can be described by the learningfunction exhibited during learning episodes that take place withina single laboratory session.

In order to explore the potential of this conjecture in the cur-rent study, we provide a descriptive analysis of the development ofchess performance in German players who start playing chess atan early age and continue with the activity for at least 10 years.Relevant for theoretical as well as practical purposes, the timecourses of expertise acquisition could thus potentially be pre-dicted. Based on the shape of the curve of improvements duringthe first years of expertise acquisition, one might be able to pre-dict the time course of improvements over the years of practice tocome (Ericsson et al., 1993; Charness et al., 2005).

METHODSDATABASEWe used archival data of the population of German play-ers recorded by the German chess federation (Deutscher

Schachbund) from 1989 to 2007. Data were kindly providedby the federation and analyzed in line with guidelines of theethics review board at Humboldt-Universität, Berlin. With over3000 rated tournaments in a year, the German chess feder-ation is one of the largest and the best-organized nationalchess federations in the world. Given that almost all Germantournaments are rated, including events such as club champi-onships, the entire playing careers of all competitive and mosthobby players in Germany are tracked in detail. This is par-ticularly important because we wanted to capture the veryfirst stage of chess skill acquisition by focusing on the veryyoung chess players who just started to play chess. The Germandatabase provides a perfect opportunity to study the initialstages of skill acquisition because even school tournaments arerecorded.

THE MEASURE—CHESS RATINGBesides precise records of players, the German federation’sdatabase and chess databases in general use an interval scale, theElo rating, for measuring skill level. Every player has an Elo ratingthat is obtained on the basis of their results against other players ofknown rating (see Elo, 1978). Average players are assumed to haverating of 1500 Elo points, experts over 2000 points, grandmaster,the best players, over 2500. Beginners usually start at around 800Elo points. The German database uses the same system but labelsthe rating as Deutsche Wertzahl (DWZ), which is highly corre-lated (r > 0.90) with the international Elo rating (Bilalic et al.,2009b).

SELECTION CRITERIA AND GROUPING OF DATA ANALYZEDThe German chess federation database contains records of over124,000 players and the average rating of these players is 1387points with standard deviation of 389 points. For all practicalpurposes, the database contains the entirety of the populationof tournament chess observations in Germany (for more infor-mation about the database, see Bilalic et al., 2009b; Vaci et al.,2014). With interest in expertise development (rather than main-tenance), we used the subset of data from all players who enteredthe database between age 6 and 20. This population consisted of1383 players that played competitive chess for at least 10 years.All players took part in tournaments in each of the 10 years. Tobe sure that the initial observation was indeed first entry intocompetitive chess, we excluded players who were already listedin the first year the federation started tracking players. For theplayers starting young, there should have been little opportunityfor expertise acquisition prior to taking part in tournaments cov-ered by the database. To track this issue, we split the sample intoage-groups (see Table 2 gender and age as well as for means andstandard deviations of games played, rating reached by year 10,change in rating between year 1 and 10, and change in rating pergame played).

Note that since we are working with the entirety of tournamentchess performances in Germany since 1989, we provide descrip-tion of the entire population of interest—chess players that playedcompetitive chess in Germany for at least 10 years (means,standard deviations, correlations that allow for an estimationof effect sizes). Generalization of findings, beyond the internal





Table 2 | Sample characteristics and summary statistics on games played and ratings.

Age N (females) Mean age (SD) Mean total number Mean rating in Mean rating change Mean rating change

of games (SD) Year 10 (SD) from Year 1–10 (SD) per games played (SD)

6–9 173 (24) 8.3 (0.9) 359.9 (196) 1705.1 (308.2) 894 (368.6) 3.21 (1.72)

10–13 689 (72) 11.6 (1.1) 239 (135, 8) 1668.8 (268.9) 640.7 (294.9) 3.52 (2.15)

14–17 414 (22) 15.2 (1.1) 201.9 (116.1) 1660 (231.6) 457.4 (235.5) 3.04 (1.9)

18–20 107 (3) 18.8 (0.8) 180 (126.7) 1582.5 (200.6) 270.0 (201.7) 1.96 (1.5)

predictions, will have to be based on replications with other orfuture databases (see e.g., Asendorpf et al., 2013).

PREDICTING AND FITTING WITH THE POWER AND EXPONENTIALFUNCTIONFits were derived with constrained optimization, requiring the Aand B parameters to take sensible values (0 < B < A < 3000)using the MATLAB Curve Fitting Toolbox. For each participantwe compared estimated and observed ratings and determinedwhether the power function or the exponential function led to asmaller squared deviation. For predictions, we only used the dataof the first 5 years to extract the parameters of the power functionand the exponential function. Then we used these parameters toextrapolate the predicted ratings for the next years (at least 5—each person in the sample had database entries for a minimumof 10 years). The predicted values were then compared to the rat-ings actually achieved. For instance, for a given participant whoplayed for 10 years, we took the performance in the first five,acting as if the trajectory data of the next years were not yet avail-able. The power function and the exponential function were fitto the data of the first 5 years in order to obtain the parame-ter values exemplified in Table 1. Next we used these values inorder to extrapolate for the coming years of tournament partici-pation. These predicted values were than compared to the actualratings obtained. For each participant we could thus compare theroot mean square error (RMSE) between power function-basedprediction and prediction based on the exponential function.

RESULTSDESCRIPTIVE STATISTICSTable 2 indicates that our sample was predominantly male.Participants starting to play tournaments younger accumulatedmore games as compared to those starting at an older age. Forinstance, the 6 to 9 year olds played twice as many games thanthe 18 to 20 year olds. The rating reached by Year 10 was similaracross age groups. Yet, this implied a much stronger improvementcompared to Year 1 for the players starting young rather than old.For instance, the youngest group showed trifold the increase ofthe oldest group. The increase in rating relative to the numberof games played was similar across age groups (with the playersstarting oldest, who showed a reduced gain per games played).

EXPONENTIAL BETTER THAN POWER FUNCTION IN PREDICTIONSAND FITSFigure 2 presents a random subset of individual time courses.Despite fluctuations from one year to the next, participantsgenerally showed increases in skill, as measured by Elo, over years

FIGURE 2 | A random sample of individual time courses.

of chess played. Some participants showed large gains especiallyin their first years. In order to systematize such observations, wetested the capability of the power function and the exponentialfunction to fit and predict the observed trajectories. Predictionis interesting for practical purposes as we can infer the skilllevel someone will have after ten years of activity based on thepattern of performance in their first years. On the other hand, pre-diction circumvents methodological problems inherent in curvefitting. For instance, one mathematical function might fit bet-ter than another, because it is flexible enough to mimic thecompetitor.

Across individuals from all age groups, the exponential func-tion provided better prediction and fit to the data than the powerfunction (Table 3). The average RMSE and its standard devi-ation were smaller for the exponential function than for thepower function (with exception of the prediction among thosestarting chess at ages 18–20). For 88% of the players, the expo-nential function was better in fitting the first 5 years the skillacquisition process, and for 62% it was better in predicting theskill level in later years. As shown in Figure 3, the distributionof RMSE values was heavily left-skewed. For a substantial pro-portion of participants neither the exponential nor the powerfunction provided an account of the dynamics of individuals’ skilldevelopment.






Table 3 | Average and standard deviation of RMSE per age group.

Prediction Fitting

Age Power Exponential Power Exponential

6–9 381.3 (179.9) 197.5 (127.4) 143.6 (88.7) 83 (37.8)

10–13 225.4 (135.6) 149.8 (99.3) 96 (62.8) 60.3 (28.1)

14–17 115.4 (84) 104.1 (63.3) 69.9 (45.9) 45.4 (20.4)

18–20 75.1 (48.1) 87.8 (56.7) 51.8 (30.5) 39 (19.6)

FIGURE 3 | Distribution of root mean square error (RMSE) of rating

points of the power function (blue) and exponential function (red) in

fitting the time course of the first 5 years (filled lines) vs. predicting

the time course for the next years based on the parameters derived

from fitting the first years (dotted lines).

INCREASING GAINS IN PARTICIPANTS STARTING YOUNGWe grouped players by the age they started to play tournamentsin order to explore reasons for the substantial problems in fitand prediction encountered with both the power and exponentialfunctions. Potentially, players starting tournament participationat older ages might have profited from substantial opportuni-ties to practice chess before they entered our window of analysis.Thus, the expected learning gains might manifest more readily inthose players that started at younger ages. Figure 4A indeed showsthat players entering tournament chess at older ages demonstratehigher skill levels by the end of their first year, while playersentering at younger ages, start at lower levels. Most notably, how-ever, the shape of the improvements deviates systematically fromthe patterns of change that would be expected based on eitherthe exponential or the power function. Both learning functionspredict that participants should show a higher absolute gain inrating points from the first to the second year compared to thegain from the second to the third year, which in turn shouldyield a higher gain as compared to the change from the thirdto the fourth year, and so on and so forth. Among the subsetof players starting young, however, the contrary seemed to bethe case (see also Figure 4B for difference values). Over the first

years of tournament participation, the absolute amount of gainper year increased rather than decreased. The deviation from theexpected learning curve might be related to year-to-year varia-tions in practice. For example, the players starting young may,at first, participate in very few tournaments, and then, in thenext few year, increase in the number of tournament games theytake part in. This is indeed the case (Figure 4C). Therefore, it isconceivable that the amount of practice which increases over theyears accounts for the dynamics of the skill increase—only oncethe players starting young take part in more and more games,their skill might start to increase in the manner predicted bythe learning curves. As we do not possess any further data onchanges in the amount of practice per year (i.e., off tournamentpractice), we cannot conclusively judge this account. However, atleast we can state that the increase in the number of tournamentgames played cannot fully account for the dynamics. As shown inFigure 4D, the change in rating per year per number of tourna-ment games played also shows an increase over the first years forplayers starting young.

FLUCTUATIONS IN GAMES PLAYED PER YEAR RELATE TO MISFIT WITHPOWER FUNCTIONIt is conceivable that the misfit and inaccurate prediction of thepower and the exponential functions are related to variability inthe number of games played per year. While we only examinedthe subset of players who played in tournaments in each of the 10years tracked, the number of games per year might have fluctu-ated. We computed the within-person (intraindividual) standarddeviation of games played per year, assuming that fits shouldbe optimal if the number of games a player takes part in doesnot change over the years. This index is equivalent to comput-ing the deviation from a zero-slope line in numbers of gamesplayed. Table 4 shows the Spearman rank order correlation ofintraindividual variability in number of games played per yearwith the RMSE obtained from fitting and predicting ratings basedon the power function and the exponential function. The corre-lations suggest that larger intraindividual variability in numberof games played per year was weakly but consistently related toworse fit in case of the power function (while the pattern was lessconsistent for the exponential function). A similar pattern wasobserved when correlating the overall number of games played toaccuracy in prediction and fitting. Participants who played moregames showed worse fits compared to participants who played lessgames. This was likely the case, because the number of gamesplayed over the ten years (a count variable) was closely linkedto the intraindividual variability in games per year (Spearmancorrelations ranging between 0.84 and 0.88 across the four agegroups).

Figure 4C suggests that variability in number of games playedper year is not purely random. Instead it can be based onan ordered pattern (inverted U-shape). Separately for each agegroup, we took the average profile in number of games played peryear (displayed in Figure 4C) as a prototypical pattern. Then, wedetermined for each participant the profile correlation betweenhis/her pattern of numbers of games played with the averagepattern of the respective age group. Our analyses suggested thatthere was substantial variability, with some participants following





Ra�ng

Games per year

Δ Ra�ng

Δ Ra�ng per gamesplayed

A

C

B

D

FIGURE 4 | (A) Charts the average time course in tournamentperformance for players starting at 6–20 years of age. (B) Shows thetime course of the gain in performance from one year to the next forthe different starting-age groups. In (C) the time course of number of

tournament games played per year is charted for the differentstarting-age groups. (D) Shows the time course of the improvement intournament performance from year to year relative to the number oftournament games played in the respective year.

the pattern represented in the group mean and others deviat-ing from it. Median within-person correlations per age groupwere r = 0.58, 0.5, 0.47, and 0.19. The percentage of individu-als showing a negative correlation with the prototypical patternwas 9.8, 15.2, 19.6, and 31.8%. However, as suggested by Table 4,the extent to which the dynamics of an individual’s number ofgames played per year was represented by the average pattern ofthe age group was not systematically related to the accuracy inpower function or exponential function fits and predictions.

OFF-THE-CURVE PATTERNS IN 2/3rds OF THE SAMPLEWe sought to provide descriptive data on the number of partic-ipants who deviated from the predictions of the learning curvesby showing smaller rather than larger rating gains during theirearly as compared to their later years of tournament participa-tion. For this we sorted individuals into tertiles based on the totalgains achieved during the first 3 years (lowest, medium, and high-est rating gains). As shown in Figure 5A, the third of players with

the lowest gains even showed small decreases in rating during thefirst years, while only the individuals with the largest gain yieldedperformance changes in line with the predictions by the learn-ing curves (i.e., larger gain per year in early rather than late years,compare Figure 5B). Players that did not improve in their firstthree tournament years caught up to some extent in later years,but did not reach the same level by year 10 as those players witha steep increase early on. Thus, irrespective of complex dynamicsof the shape of the performance curve, the first years do seem tooffer a proxy for predicting the level a player will eventually reach.

COHORT DIFFERENCESThere have been many changes in resources available for chessplayers since 1989. We analyzed the time course in developmentof chess ratings separately for different cohorts in order to explorewhether deviations from the pattern predicted by the learningcurves varied in relation to the historical period that a chesscareer was started. Deviations from the learning curve were not






Table 4 | Spearman rank order correlations of (a) indices of regularity

in numbers of games played per year with (b) the fitting and

prediction error.

Age Prediction Fitting

Power Exponential Power Exponential

SD of games peryear

6–9 0.13 −0.22 0.29 −0.0510–13 0.14 −0.16 0.45 0.20

14–17 0.22 0.02 0.36 0.24

18–20 0.18 0.20 0.47 0.44

Overall number ofgames

6–9 0.22 −0.30 0.37 −0.0110–13 0.21 −0.14 0.50 0.21

14–17 0.24 −0.02 0.39 0.22

18–20 0.18 0.23 0.56 0.48

Prototypical profilein games per year

6–9 −0.03 −0.36 0.21 −0.0210–13 −0.06 −0.17 0.06 −0.06

14–17 0.04 −0.08 0.08 0.05

18–20 0.22 −0.01 0.23 0.22

Years played

Δ Ra�ng

Ra�ng

A

BYears played

FIGURE 5 | (A) Charts the average time course in tournament performancefor players starting at 6–20 years of age grouped by improvement over thefirst three years. (B) Shows the change in ELO per year.

Years played

FIGURE 6 | Average time course in tournament performance for

players starting at 6–20 years of age grouped by birth cohort

(1970–1990). Due to small n (11) we dropped the 14–17 year old startersborn between 1970 and 1975.

accounted for by cohort. Rather, for all 5-year cohorts from 1970to 1990 and age-groups displayed in Figure 6, the increase in rat-ing during the first years of performance was linear or positivelyaccelerated. The pattern of negative acceleration (larger gains inearlier as compared to later years, compatible with the learningcurves) was not observed.

Age-groups and cohorts differed more with respect to the rat-ing level they started out with (i.e., reached by end of their firstyear) than with respect to their level of performance in Year10. As already observable in Figure 5A, people starting to playtournaments at younger age, started out at a lower level. In addi-tion, Figure 6 shows that later cohorts started at lower levels.This might be taken to suggest that players starting young in latecohorts are the best candidates to track trajectories in chess per-formance based on tournament ratings, while ratings of playersstarting older and earlier cohorts might be shaped more stronglyby off-tournament practice.

GAIN IN RATING FROM GAMES PLAYEDThe above analyses suggest that the success of the power func-tion and the exponential function in predicting development ofchess performance might be rather limited due to quantitativeand qualitative misfit. Furthermore, the number of tournamentgames played seemed to be linked to deviations from the learn-ing curve. Therefore, we sought to describe the extent to whichearly vs. late years in playing tournament chess are related togain in rating as well as performance level reached by Year 10.For this we used games played per year and gain in rating peryear. We applied Spearman rank order correlations separately foreach age group and year of tournament participation. Figure 7A





-.20-.10.00.10.20.30.40.50.60.70

2 3 4 5 6 7 8 9 10

Spearman correla�on

Games played in year correlated with Δ ra�ng of year

-.40

-.30

-.20

-.10

.00

.10

.20

2 3 4 5 6 7 8 9 10

Correla�on of games per year with gain per games played in each year

-.20-.10.00.10.20.30.40.50.60.70

2 3 4 5 6 7 8 9 10

Gain from Year 1 to 10 correlated with Δ ra�ng of year per games played in year

-.20-.10.00.10.20.30.40.50.60.70

2 3 4 5 6 7 8 9 10

Gain from Year 1 to 10 correlated with Δ ra�ng of year

6-9

10-13

14-17

18-20

Star�ng-age

Years played

A

C D

Years played

Years played

B

Years played

FIGURE 7 | Spearman rank order correlations per age group between performance variables taken from single years and rating gain per year (A,B) or

overall rating gain (C,D).

shows that number of games played per year is related to between-person differences in gain in rating. Participants playing moretournament games in a year tend to show a larger increase in rat-ing compared to those playing less games. This holds consistentlyacross age-groups and especially so for early years of tournamentparticipation. However, diminishing returns seem to be observ-able with respect to the extent to which more tournament gamescan lead to an increase in rating. Figure 7B shows that the rela-tionship between (a) games played per year and (b) gain pergames played per year can become negative. Thus, overall it doesnot seem to be the case that playing more tournament games canlead to an increase in efficiency in taking gains in rating froma tournament game. For instance, those players starting tourna-ment participation at age 10–13 who played more tournamentgames, seemed to show a reduced gain in rating per tournamentgame played in their middle years.

The gain in rating that players show from Year 1 to Year 10can be predicted by gain in rating per year in early years oftournament participation. As depicted in Figure 7C, gain in later

years is less predictive of the overall increase in rating. While thepower and the exponential function would have predicted that wecan observe large gains in rating in early years, we thus, somewhatanalogously, observer a larger predictive power of between-persondifferences in early as compared to late years of chess tournamentparticipation. Apart from the gain per year, also the gain per yearrelative to the number of games played per year could be usedto predict the overall increase in rating between Year 1 and 10(Figure 7D). Participants who, during the first years of tourna-ment participation, efficiently increased their rating per gamesplayed, ended up at a higher performance level than those, whodid not show a large gain per games played during early years.

SELECTIVE ATTRITIONFinally, we checked for selective attrition. While in our mainanalyses we only used 10 years of subsequent tournament partic-ipation, some participants provided records for additional years(up to 19 years overall). Rank order correlations indicated thatthe number of overall years of tournament participation per age






group was neither systematically related to gain between Year 1and Year 10, nor the gain in rating within the first 3 years (rsbetween −0.10 and 0.16).

DISCUSSIONIn the current work we have explored the potential of the power-and the negative exponential learning functions to account for thedevelopment of chess performance measured in ratings based ontournament outcomes. In line with re-evaluations of the powerlaw of practice (Heathcote et al., 2000; Haider and Frensch, 2002),we documented that the exponential function was better than thepower function in fitting and predicting the time course of chessratings over years of practice. However, a crucial aspect shared byboth of these mathematical functions and the underlying theoriesof skill acquisition was not reflected in the data. While accord-ing to the power- as well as the negative exponential functionplayers should achieve large absolute gains early in practice andsmall gains later in practice, this was not the case for many ofthe participants. Rather, many players started to show substantialimprovements only after their first years of tournament participa-tion. They were playing off the learning curves suggested by skillacquisition theories. If expertise acquisition is not well describedby learning functions used to describe skill acquisition, the link-ing of underlying cognitive processes of attention and learningthat proceed on time-scales measuring milliseconds to hours withlearning processes that proceed on time-scales measured in yearsseems much less straightforward than one could have hoped for(i.e., Anderson, 2002).

Many players showed an acceleration of gain in rating in thefirst years of tournament participation, followed by a decelera-tion. Based on the power function and the exponential functionwe would have expected to only find the latter. Newell et al.(2001) suggested to mathematically and conceptually accommo-date such findings by assuming a mixture of learning processestaking place on different time scales. Acceleration followed bydeceleration could be captured by a sigmoid function that con-sists of two exponential components, a positive (acceleration)and a negative one (deceleration). Learning opportunities andefficiency in using them might increase during first years of tour-nament performance for many players, while in later years returnsof investing in chess performance are diminishing. In line withthis view, year-long trajectories of skill acquisition might be betterunderstood from a perspective that takes lifespan-developmentaland educational changes into account (Li and Freund, 2005).For instance, players starting to take part in tournaments at ayoung age are likely to promote changes in self-regulation strate-gies available (Lerner et al., 2001; Freund and Baltes, 2002) andacquire the potential to shape their social and learning environ-ment. Their ability to learn about chess from (foreign language)media and options to travel to and communicate with other play-ers will increase. Deliberate practice (cf. Ericsson et al., 1993)might require that young players develop skills to competentlyuse of their motivational resources, by, for instance, schedulingwork on skill acquisition such that as many of the activities aspossible are intrinsically motivating (cf. Rheinberg and Engeser,2010 as well as Christophel et al., 2014, for training of motiva-tional competence). Underlining this challenge, Coughlan et al.,

2014 reported that participants in the expert group of their studyrated their practice as more effortful and less enjoyable comparedto other participants. The experts were successful in improvingperformance, by predominantly practicing the skill they wereweaker at. However, such gains in potential to learn might formany players no longer compensate for the physical and socialchanges faced during puberty (Marceau et al., 2011; Hollensteinand Lougheed, 2013), at the end of adolescence, during sec-ondary education, family formation or labor force participation.Future research should thus try to simultaneously account fordevelopment in the individual, the opportunities provided by theenvironment (cf. Ram et al., 2013) and to model different trajec-tories in one framework (e.g., Grimm et al., 2010; Ram and Diehl,in press).

For skill acquisition mechanisms such as chunking, negativeexponential learning can be explained by a reduction in learningopportunities (cf. Heathcote et al., 2000). The later in practice, thefewer chunks are yet to be learned. While a deceleration of learn-ing should be observed late in practice, such an account does notpreclude that strong increases in learning opportunities early inpractice can lead to an acceleration of chunks acquired per timeinvested. It appears that, for at least some players, opportunitiesand efficiency in increasing chess performance are already fullypresent at the time they start to play tournaments. They start atthe turning point of the sigmoid function. The “upper” negativeexponential portion of the sigmoid is sufficient to describe theirperformance gains, which are large in their early years and thendiminish as performance approaches the asymptote. For otherplayers, both positive and negative exponential portions of thesigmoid function are needed to represent the dynamics of theirchess performance over time. These players appear to be less sat-urated with respect to learning opportunities and efficiency whenstarting to take part in tournaments covered by the database. Theythus first show an acceleration in rating gains per year, followedby the deceleration when approaching asymptote.

In line with these speculations, Howard (2014) reportedan average trajectory of rating increases showing decelerationonly for International Chess Federation (FIDE) players (ratherthan acceleration followed by deceleration). The shape of thecurve reported by Howard matches the exponential curve fromFigure 1A. Starting at an average of about 2200 points, the sam-ple mean increased beyond 2500 points with practice. Differentfrom the database used in the current study, the threshold to belisted in the FIDE database is high (cf. Vaci et al., 2014 for adiscussion of problems implied by restriction of range in chessdatabases). Likely, players were already taking full advantage ofopportunities to improve chess performance when entering thedatabase so that an acceleration in rating gain with practice wasno longer possible. Descriptive analyses suggest that the dynamicin rating improvement that players at the international level showwith practice seems consistent with the negatively acceleratedexponential function. As implied by the exponential function, therelative learning rate (RLR) estimated based on the average datapublished by Howard (2014) is constant. While the power func-tion should lead to a decrease of RLR with practice (cf. Heathcoteet al., 2000), the RLR is fluctuating around 20%. Focusing onthe first half of practice in order to avoid inflation of RLR at the





end of the practice curve, we obtained an r = 0.11 correlationof RLR with time point. Thus there was no hint toward adecrease.

Our correlational analyses suggest that interindividual vari-ability in rating gain over the course of ten years of tournamentparticipation can be predicted by between-person differences inperformance during the first years. Even by taking data from sin-gle years, number of games played, rating points gained or ratingpoints gained per games played, allow to predict overall gain ata moderate level. While the power and the exponential learn-ing curve would suggest that the first years of practice shouldbe important because of the large performance gains, we thuscan somewhat analogously conclude that the first years are moreimportant than later years for predicting between-person differ-ences in performance level reached on the long run (cf. Ackermanand Woltz, 1994).

We focused on examining changes in rating with year of prac-tice (rather than number of games played, cf. Howard, 2014).This allowed us to explore changes in rating gain and rating gainper games played with age and cohort. Yet, a direct compari-son of the capability to capture performance change is lackingso far for the two potential time scales, (1) number of gamesplayed, (2) chronological time in years, as well as (3) a mixtureof both scales. Several issues are worth considering when explor-ing the complexity of models needed to account for expertiseacquisition over years, as compared to models of skill acquisi-tion in hour-long laboratory sessions. In the lab, quantity andquality of practice per unit of time is usually well controlled. Inskill acquisition processes outside the lab they might vary con-siderably over the years of practice an individual engages in. Inaddition, potential cohort differences should not be neglected (cf.Gobet et al., 2002; van Harreveld et al., 2007; Connors et al.,2011). Future work should consider how data on both, quan-tity of practice and quality of practice, can be used to explainthe time course of chess skill development (cf. Baker et al., 2003;Charness et al., 2005; Gobet and Campitelli, 2007; Howard, 2014).Apart from obtaining data on the amount of off-tournamentlearning opportunities, available data sets could be used to gaugevariability in specific aspects of the learning opportunities. Forinstance, taking part in tournaments with large spread in oppo-nent strength might provide more opportunities for improve-ment as compared to tournaments with more homogenouscompetitors.

REFERENCESAckerman, P. L., and Woltz, D. J. (1994). Determinants of learning and perfor-

mance in an associative memory/substitution task: task constraints, individualdifferences, and volition. J. Educ. Psychol. 86, 487–515. doi: 10.1037/0022-0663.86.4.487

Anderson, J. R. (1982). Acquisition of cognitive skill. Psychol. Rev. 89, 369–406. doi:10.1037/0033-295X.89.4.369

Anderson, J. R. (2002). Spanning seven orders of magnitude: a challengefor cognitive modeling. Cogn. Sci. 26, 85–112. doi: 10.1207/s15516709cog2601_3

Asendorpf, J. B., Conner, M., De Fruyt, F., De Houwer, J., Denissen, J. J., Fiedler, K.,et al. (2013). Recommendations for increasing replicability in psychology. Eur.J. Pers. 27, 108–119. doi: 10.1002/per.1919

Baker, J., Côté, J., and Abernethy, B. (2003). Sport-specific practice and the devel-opment of expert decision-making in team ball sports. J. Appl. Sport Psychol. 15,12–25. doi: 10.1080/10413200305400


Bilalic, M., McLeod, P., and Gobet, F. (2010). The mechanism of the Einstellung(Set) Effect: a pervasive source of cognitive bias. Curr. Dir. Psychol. Sci. 19,111–115. doi: 10.1177/0963721410363571

Bilalic, M., McLeod, P., and Gobet, F. (2008a). Inflexibility of experts–reality ormyth? Quantifying the Einstellung effect in chess masters. Cogn. Psychol. 56,73–102. doi: 10.1016/j.cogpsych.2007.02.001


Bilalic, M., McLeod, P., and Gobet, F. (2009a). Specialization effect and its influ-ence on memory and problem solving in expert chess players. Cogn. Sci. 33,1117–1143. doi: 10.1111/j.1551-6709.2009.01030.x

Bilalic, M., Smallbone, K., McLeod, P., and Gobet, F. (2009b). Why are (the best)women so good at chess? Participation rates and gender differences in intel-lectual domains. Proc. R. Soc. B Biol. Sci. 276, 1161–1165. doi: 10.1098/rspb.2008.1576

Campitelli, G., Gobet, F., Head, K., Buckley, M., and Parker, A. (2007). Brain local-ization of memory chunks in chessplayers. Int. J. Neurosci. 117, 1641–1659. doi:10.1080/00207450601041955

Campitelli, G., Gobet, F., and Parker, A. (2005). Structure and stimulus famil-iarity: a study of memory in chess-players with functional magnetic res-onance imaging. Span. J. Psychol. 8, 238–245. doi: 10.1017/S1138741600005126

Charness, N., Reingold, E. M., Pomplun, M., and Stampe, D. M. (2001). The per-ceptual aspect of skilled performance in chess: evidence from eye movements.Mem. Cogn. 29, 1146–1152. doi: 10.3758/BF03206384

Charness, N., Tuffiash, M., Krampe, R., Reingold, E., and Vasyukova, E. (2005). Therole of deliberate practice in chess expertise. Appl. Cogn. Psychol. 19, 151–165.doi: 10.1002/acp.1106

Chase, W. G., and Simon, H. A. (1973). Perception in chess. Cognit. Psychol. 4,55–81. doi: 10.1016/0010-0285(73)90004-2

Christophel, E., Gaschler, R., and Schnotz, W. (2014). Teachers’ expertise in feed-back application adapted to the phases of the learning process. Front. Psychol.5:858. doi: 10.3389/fpsyg.2014.00858

Connors, M. H., Burns, B. D., and Campitelli, G. (2011). Expertise in complex deci-sion making: the role of search in chess 70 years after de Groot. Cogn. Sci. 35,1567–1579. doi: 10.1111/j.1551-6709.2011.01196.x

Connors, M. H., and Campitelli, G. (2014). Expertise and the representation ofspace. Front. Psychol. 5:270. doi: 10.3389/fpsyg.2014.00270

Coughlan, E. K., Williams, A. M., McRobert, A. P., and Ford, P. R. (2014). Howexperts practice: a novel test of deliberate practice theory. J. Exp. Psychol. Learn.Mem. Cogn. 40, 449–458. doi: 10.1037/a0034302

Crossman, E. R. F. W. (1959). A theory of the acquisition of speed-skill. Ergonomics2, 153–166. doi: 10.1080/00140135908930419

Elo, A. E. (1978). The Rating of Chess Players, Past and Present. New York, NY: Arco.Ericsson, K. A., Krampe, R. T., and Tesch-Roemer, C. (1993). The role of deliberate

practice in the acquisition of expert performance. Psychol. Rev. 100, 363–406.doi: 10.1037/0033-295X.100.3.363

Freund, A. M., and Baltes, P. B. (2002). Life-management strategies of selec-tion, optimization, and compensation: measurement by self-report and con-struct validity. J. Pers. Soc. Psychol. 82, 642–662. doi: 10.1037/0022-3514.82.4.642

Gaschler, R., and Frensch, P. A. (2007). Is information reduction an item-specificor an item-general process?. Int. J. Psychol. 42, 218–228. doi: 10.1080/00207590701396526

Gobet, F., and Campitelli, G. (2007). The role of domain-specific practice, hand-edness and starting age in chess. Dev. Psychol. 43, 159–172. doi: 10.1037/0012-1649.43.1.159

Gobet, F., Campitelli, G., and Waters, A. J. (2002). Rise of human intelligence:comments on Howard (1999). Intelligence 30, 303–311. doi: 10.1016/S0160-2896(02)00083-1

Gobet, F., and Simon, H. A. (1996a). Recall of rapidly presented random chesspositions is a function of skill. Psychon. Bull. Rev. 3, 159–163. doi: 10.3758/BF03212414

Gobet, F., and Simon, H. A. (1996b). Recall of random and distorted posi-tions: implications for the theory of expertise. Mem. Cognit. 24, 493–503. doi:10.3758/BF03200937






Gobet, F., and Simon, H. A. (1996c). Templates in chess memory: a mechanism forrecalling several boards. Cogn. Psychol. 31, 1–40. doi: 10.1006/cogp.1996.0011

Gobet, F., and Simon, H. A. (1998). Expert chess memory: revisiting the chunkinghypothesis. Memory 6, 225–255. doi: 10.1080/741942359

Grimm, K. J., Ram, N., and Estabrook, R. (2010). Growth mixture models withinherently nonlinear change models using Mplus and OpenMx. MultivariateBehav. Res. 45, 887–909. doi: 10.1080/00273171.2010.531230

Haider, H., and Frensch, P. A. (2002). Why aggregated learning follows the powerlaw of practice when individual learning does not: comment on Rickard (1997,1999), Delaney et al. (1998), and Palmeri (1999). J. Exp. Psychol. Learn. Mem.Cogn. 28, 392–406. doi: 10.1037/0278-7393.28.2.392

Heathcote, A., Brown, S., and Mewhort, D. J. K. (2000). The power law repealed:the case for an exponential law of practice. Psychon. Bull. Rev. 7, 185–207. doi:10.3758/BF03212979

Hollenstein, T., and Lougheed, J. P. (2013). Beyond storm and stress: typicality,transactions, timing, and temperament to account for adolescent change. Am.Psychol. 68, 444–454. doi: 10.1037/a0033586

Howard, R. W. (2014). Learning curves in highly skilled chess players: a test ofthe generality of the power law of practice. Acta Psychol. 151, 16–23. doi:10.1016/j.actpsy.2014.05.013

Kiesel, A., Kunde, W., Pohl, C., Berner, M. P., and Hoffmann, J. (2009). Playingchess unconsciously. J. Exp. Psychol. Learn. Mem. Cogn. 35, 292–298. doi:10.1037/a0014499

Kramer, A. F., Strayer, D. L., and Buckley, J. (1991). Task versus component consis-tency in the development of automatic processing: a psychophysiological assess-ment. Psychophysiology 28, 425–437. doi: 10.1111/j.1469-8986.1991.tb00726.x

Lee, F. J., and Anderson, J. R. (2001). Does learning a complex task have to becomplex?: a study in learning decomposition. Cogn. Psychol. 42, 267–316. doi:10.1006/cogp.2000.0747

Leone, M. J., Slezak, D. F., Cecchi, G. A., and Sigman, M. (2014). The geometry ofexpertise. Front. Psychol. 5:47. doi: 10.3389/fpsyg.2014.00047

Lerner, R. M., Freund, A. M., De Stefanis, I., and Habermas, T. (2001).Understanding developmental regulation in adolescence: the use of the selec-tion, optimization, and compensation model. Hum. Dev. 44, 29–50. doi: 10.1159/000057039

Li, S.C., and Freund, A. M. (2005). Advances in lifespan psychology: a focuson biocultural and personal influences. Res. Hum. Dev. 2, 1–23. doi:10.1080/15427609.2005.9683342

Logan, G. D. (1988). Toward an instance theory of automatization. Psychol. Rev. 95,492–527. doi: 10.1037/0033-295X.95.4.492

Logan, G. D. (1992). Shapes of reaction-time distributions and shapes of learningcurves: a test of the instance theory of automaticity. J. Exp. Psychol. Learn. Mem.Cogn. 18, 883–914. doi: 10.1037/0278-7393.18.5.883

Marceau, K., Ram, N., Houts, R. M., Grimm, K. J., and Susman, E. J. (2011).Individual differences in boys’ and girls’ timing and tempo of puberty: mod-eling development with nonlinear growth models. Dev. Psychol. 47, 1389–1409.doi: 10.1037/a0023838

Marewski, J. N., and Olsson, H. (2009). Beyond the null ritual: formal modelingof psychological processes. Zeitschrift für Psychologie/J. Psychol. 217, 49–60. doi:10.1027/0044-3409.217.1.49

Newell, A., and Rosenbloom, P. S. (1981). “Mechanisms of skill acquisition andthe law of practice,” in Cognitive Skills and Their Acquisition, Vol. 1, ed J. R.Anderson (Hillsdale, NJ: Erlbaum), 1–55.

Newell, K. M., Liu, Y.T., and Mayer-Kress, G. (2001). Time scales in motor learningand development. Psychol. Rev. 108, 57–82. doi: 10.1037/0033-295X.108.1.57

Pitt, M. A., Myung, I. J., and Zhang, S. (2002). Toward a method for selectingamong computational models for cognition. Psychol. Rev. 109, 472–491. doi:10.1037/0033-295X.109.3.472

Ram, N., Coccia, M., Conroy, D., Lorek, A., Orland, B., Pincus, A., et al.(2013). Behavioral landscapes and change in behavioral landscapes: a multi-ple time-scale density distribution approach. Res. Hum. Dev. 10, 88–110. doi:10.1080/15427609.2013.760262

Ram, N., and Diehl, M. (in press). “Multiple time-scale design and analysis:pushing towards real-time modeling of complex developmental processes,” inHandbook of Intraindividual Variability Across the Lifespan, eds M. Diehl, K.Hooker, and M. Sliwinski (New York, NY: Routledge). Available online at: http://www.routledge.com/books/details/9780203113066/

Reingold, E. M., Charness, N., Pomplun, M., and Stampe, D. M. (2001). Visual spanin expert chess players: evidence from eye movements. Psychol. Sci. 12, 48–55.doi: 10.1111/1467-9280.00309

Reingold, E. M., and Sheridan, H. (2011). “Eye movements and visual exper-tise in chess and medicine,” in Oxford Handbook on Eye Movements, edsS. P. Liversedge, I. D. Gilchrist, and S. Everling (Oxford University Press),528–550.

Rheinberg, F., and Engeser, S. (2010). “Motive training and motivational compe-tence,” in Implicit Motives, eds O. C. Schultheiss and J. C. Brunstein (Oxford:Oxford University Press), 510–548. doi: 10.1093/acprof:oso/9780195335156.003.0018

Roberts, S., and Pashler, H. (2000). How persuasive is a good fit? A com-ment on theory testing. Psychol. Rev. 107, 358–367. doi: 10.1037/0033-295X.107.2.358

Vaci, N., Gula, B., and Bilalic, M. (2014). Restricting range restricts conclusions.Front. Psychol. 5:569. doi: 10.3389/fpsyg.2014.00569

van Harreveld, F., Wagenmakers, E.-J., and van der Maas, H. L. J. (2007). The effectsof time pressure on chess skill: an investigation into fast and slow processesunderlying expert performance. Psychol. Res. 71, 591–597. doi: 10.1007/s00426-006-0076-0

Wagenmakers, E. J. (2003). How many parameters does it take to fitan elephant? J. Math. Psychol. 47, 580–586. doi: 10.1016/S0022-2496(03)00064-6

Waters, A. J., Gobet, F., and Leyden, G. (2002). Visuospatial abilities of chessplayers. Br. J. Psychol. 93, 557–565. doi: 10.1348/000712602761381402


Received: 31 May 2014; accepted: 02 August 2014; published online: 22 August 2014.Citation: Gaschler R, Progscha J, Smallbone K, Ram N and Bilalic M (2014) Playingoff the curve - testing quantitative predictions of skill acquisition theories in develop-ment of chess performance. Front. Psychol. 5:923. doi: 10.3389/fpsyg.2014.00923This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Gaschler, Progscha, Smallbone, Ram and Bilalic. This is anopen-access article distributed under the terms of the Creative Commons AttributionLicense (CC BY). The use, distribution or reproduction in other forums is permit-ted, provided the original author(s) or licensor are credited and that the originalpublication in this journal is cited, in accordance with accepted academic practice.No use, distribution or reproduction is permitted which does not comply with theseterms.


http://www.routledge.com/books/details/9780203113066/

http://www.routledge.com/books/details/9780203113066/











OPINION ARTICLEpublished: 14 August 2014

doi: 10.3389/fpsyg.2014.00878

Checkmate to deliberate practice: the case of MagnusCarlsenFernand Gobet1* and Morgan H. Ereku2

1 Department of Psychological Sciences, University of Liverpool, Liverpool, UK2 Department of Psychology, Brunel University, Uxbridge, UK*Correspondence: [email protected]


Michael H. Connors, Macquarie University, Australia

Keywords: chess, deliberate practice, expertise, superior performance, talent

The role of practice in the acquisition ofexpertise has been a key research questionat least since Bryan and Harter’s (1899)study on expertise in Morse telegraphy,which proposed that it takes 10 yearsto become an expert. The framework ofdeliberate practice (Ericsson et al., 1993)has taken an extreme position by deny-ing the role of talent in most domains andstating that superior performance is anincreasing monotonic function of delib-erate practice—the more goal-orientedpractice, the higher the level of skill. Forexample, Ericsson et al. (1993) argue that“individual differences in ultimate perfor-mance can largely be accounted for bydifferential amounts of past and currentlevels of practice” (p. 392). The deliber-ate practice framework has captured theimagination of the popular press, as canbe seen by the publication of several pop-science books such as Talent is Overrated(Colvin, 2008), Outliers (Gladwell, 2008)and Bounce (Syed, 2011).

In recent years, this framework hasbeen criticized in academic circles; forexample, in retrospective studies, theamount of deliberate practice accountsfor only about one third of the vari-ance in expertise in music and in chess(Hambrick et al., 2014). More naturalis-tic data also question the validity of theframework. As top performers have spentsimilar number of hours to improve andmaintain their skills, the fact that indi-viduals such as Roger Federer in ten-nis, Michael Jordan in basketball, UsainBolt in sprint or Michael Schumacherin auto racing have so outrageouslydominated their sport throws consider-able doubt on the deliberate practiceframework.

A particularly spectacular example isprovided by chess grandmaster MagnusCarlsen (Norway), who became worldchampion in classic chess in November2013 by beating Viswanathan Anand(India) and who also became worldchampion in rapid chess (15 min +10 s additional time per move) andspeed chess (3 min + 2 s additionaltime per move) in June 2014. In theJune 2014 rating list published by theWorld Chess Federation (http://ratings.fide.com/toparc.phtml?cod=309), 23-yearold Carlsen is ranked first with 2881points1. This is just one point below 2882,the highest rating in chess history thatCarlsen held in May 2014. There is a 66-point difference between him and the sec-ond player, grandmaster Levon Aronian(Armenia, 2815 points; see Table 1). Thisdifference is nearly the same as thatbetween the 2nd and the 14th player inthe list (63 points), Dutch grandmasterAnish Giri (2752 points). Table 1 showsthe rating of Carlsen and of the ten play-ers following him in the list. A one-samplet-test confirms that Carlsen’s rating is sta-tistically different from the next ten grand-masters (M = 2780.6), t(9) = −19.38, p <

0.001, mean difference = −100.4; 95% CI[−112.1, −88.7]. One hundred points is aconsiderable difference: it is half a standarddeviation in skill and means that, against

1 To measure chess players’ skill level, the World ChessFederation (FIDE) uses the rating scale developed byElo (1978), which is an interval scale that computesplayers’ rating as a function of their results againstother players of known rating. The scale has a nor-mal distribution with a theoretical mean of 1500 anda standard deviation of 200 points. Grand Masters aretypically rated above 2500 points. The best players ofthe world have around 2800 points and the weakestplayers less than 1200 points.

the very best players in the world, Carlsen’sprobability of winning is 63.7%.

To test the monotonic assumption, wecollected information from the internetand biographies about the age at whichthese grandmasters started playing chessand about their current age (see Table 1).Starting age is a good approximation ofwhen players started practicing seriously(i.e., using some form of deliberate prac-tice), as most of these players obtainedoutstanding results in youth competitionsa few years after starting playing chess,and indeed obtained the grandmaster titlerapidly. In the case of Carlsen, he has statedthat he had learned the rules at 5 yearsbut started practicing seriously only at8 years (see Gobet and Campitelli, 2007)2.To be consistent, we used starting ageanyway. (Note that this bias adds yearsof deliberate practice, and thus is in afavor of the monotonic assumption.) If themonotonic assumption is correct, Carlsenshould have accumulated more hours ofdeliberate practice than the other play-ers, given the way he dominates the chessworld. We did find that Carlsen’s num-ber of years of deliberate practice (18years) is different to the average of thefollowing ten best players in the world(M = 24.6 years), t(9) = 2.83, p < 0.05,mean difference = 6.6 years; 95% CI [1.33,11.87]. However, this result is exactly theopposite of what is predicted by deliber-ate practice: on average, Carlsen practicedstatistically significantly fewer years thanthe other players. (Note also that, for theplayers in Table 1, the correlation between

2 Ericsson et al. (2007) explanation that prodigies’ highlevels of performance can be accounted for by theamount of deliberate practice made possible by a veryearly start does not apply in Carlsen’s case.











http://ratings.fide.com/toparc.phtml?cod=309




Gobet and Ereku Checkmate to deliberate practice

Table 1 | Rank, country, rating, starting age, current age, and number of years of practice of the 11 top players in the world (June 2014).

Rank Name Country Rating Starting age Current age Number of years of practice

1 Carlsen, Magnus Norway 2881 5 23 18

2 Aronian, Levon Armenia 2815 9 31 22

3 Grischuk, Alexander Russia 2792 4 30 26

4 Caruana, Fabiano Italy 2791 5 21 16

5 Anand, Viswanathan India 2785 6 44 38

6 Kramnik, Vladimir Russia 2783 5 38 33

7 Nakamura, Hikaru USA 2775 7 26 19

8 Topalov, Veselin Bulgaria 2772 7 39 32

9 Karjakin, Sergey Russia 2771 5 24 19

10 Vachier-Lagrave, Maxime France 2762 4 23 19

11 Dominguez Perez, Leinier Cuba 2760 8 30 22

Source: http:// ratings.fide.com/ toparc.phtml?cod=309.

rating and the number of years of practiceis negative (r = −0.21) but not statisticallysignificant (p = 0.55)).

In this analysis, we have assumed that,at the top level, all players practice withextreme dedication and with the besttraining methods available. If expertisewas solely a monotonic function of prac-tice, then it follows that Carlsen, wholearned the rules at age of five but startedplaying chess seriously at the relatively oldage of eight, should be much weaker thanmost of the ten players that follow himin the international rating list, as theseopponents had time to clock in substan-tially more deliberate practice (on aver-age, at least 6.6 years more). The fact thatCarlsen dominates the chess world so out-rageously, being world champion not onlyin classic chess but also in rapid chess andin blitz, refutes this hypothesis, central tothe theory of deliberate practice.

Several objections can be leveled atthis analysis. We discuss three of them,and show that they do not invalidate ourargument. First, Carlsen’s prodigious skillthroughout adolescence and early adult-hood may not be as remarkable as it firstappears, as numerous young players per-form better that their older competitors.For example, Howard (1999) has shownthat the top chess players are increasinglyyounger. Key changes have taken place inthe last decades that enable more efficientpractice (Gobet et al., 2002). In partic-ular, the quality and quantity of chessbooks have dramatically increased overthe last decades, and chess programs andcomputer databases have revolutionizedtraining methods. That more efficient

deliberate practice should lead to quickerprogress is consistent with Ericsson et al.’s(1993) framework. However, as all play-ers in the Table have benefitted from theseimprovements in training, this factor doesnot explain away Carlsen’s superiority.

Second, it could be argued that, justlike in sport, age plays an important rolein chess and youth will give an edgeto younger top competitors. It is knownthat the effects of ageing occur depress-ingly early with cognitive variables suchas reasoning, visualization and processingspeed, peak performance being observedin the early to mid-twenties (Salthouse,2009). However, whether this is a key fac-tor in chess is unclear, as six of the absolutetop players shown in Table 1 are 30 yearsold or older. In addition, Gary Kasparovand Viswanathan Anand were still worldchampions when they were 37 and 44 yearsold, respectively. In any case, in Table 1the correlation between age and rating(r = −0.21) is not statistically significant(p = 0.54), but Carlsen is reliably youngerthan the other ten top players, t(9) = 3.16,p < 0.05, mean difference = 7.6 years;95% CI [2.16, 13.04]. Nevertheless, the agevariable does not explain why Carlson is soclearly better than the four players who areroughly his age.

Third, Carlsen might have engaged inmore intense deliberate practice. Althoughwe do not know the details of Carlsen’straining, this is unlikely, in particular ifwe use Ericsson et al.’s (1993) criterionthat deliberate practice is not enjoyable.In a recent interview, Carlsen said that“in chess training, I do the things I enjoy.I don’t particularly enjoy playing against

computers, so I don’t do that” (Anders,2014). In addition, he is a keen sportsman,with a penchant for playing or watchingfootball rather than practicing chess inten-sively (Sujatha, 2013).

Thus, the question arises, in the riskof offending the proponents of deliber-ate practice: Does Carlsen have a par-ticular talent for chess? The answer tothis question is so obvious in the chessworld that it is not even posed—Carlsenis known as the “Mozart of chess.” Severalfactors support the hypothesis of talent.Carlsen showed clear signs of intellectualprecocity early in his life. At the age offive, he knew “the area, population, flag,and capital of all the countries of theworld,” and memorized similar informa-tion for all Norway’s 430 municipalities(Agdestein, 2004, p. 10). He became agrandmaster just five years after startingplaying chess seriously, at the age of 13years and 148 days3. He has also adopted ahighly unconventional approach to chess.While most grandmasters specialize inspecific openings that they study at greatlength (Chassy and Gobet, 2011), oftenusing computers, Carlsen plays a widerange of openings and avoids known vari-ations, even accepting inferior positionsas a consequence of this choice. Ratherthan preparing lengthy opening varia-tions, he relies on his uncanny ability tofind near-optimal moves in middle gamesand endgames. Together with scientificresearch, the case of Magnus Carlsen

3 This contradicts another key prediction of the delib-erate practice framework: “More specifically, expertperformance is not reached with less than 10 years ofdeliberate practice” (Ericsson et al., 1993, p. 372).






Gobet and Ereku Checkmate to deliberate practice

demonstrates that deliberate practice isnecessary, but not sufficient, for achiev-ing high levels of expert performance(Campitelli and Gobet, 2011).

ACKNOWLEDGMENTWe thank the reviewer for usefulcomments.

REFERENCESAgdestein, S. (2004). Wonderboy: How Magnus Carlsen

became the Youngest Chess Grandmaster in theWorld. The Story and the Games. Alkmaar: New InChess.

Anders, G. (2014). Chess Champ Carlsen delightsSilicon Valley with dry wit. Available online at: http://www.forbes.com/sites/georgeanders/2014/01/17/chess-champ-carlsen-delights-silicon-valley-with-dry-wit/ (Accessed 7 July, 2014).

Bryan, W. L., and Harter, N. (1899). Studies on thetelegraphic language. The acquisition of a hierar-chy of habits. Psychol. Rev. 6, 345–375.

Campitelli, G., and Gobet, F. (2011). Deliberatepractice: necessary but not sufficient. Curr. Dir.Psychol. Sci. 20, 280–285. doi: 10.1177/0963721411421922

Chassy, P., and Gobet, F. (2011). Measuringchess experts’ single-use sequence knowl-edge: an archival study of departure from‘theoretical’ openings. PLoS ONE 6:e26692. doi:10.1371/journal.pone.0026692

Colvin, G. (2008). Talent is overrated. What ReallySeparates World-Class Performers from EverybodyElse. New York, NY: Penguin.

Elo, A. (1978). The Rating of Chessplayers, Past andPresent. New York, NY: Arco.

Ericsson, K. A., Krampe, R. T., and Tesch-Römer, C.(1993). The role of deliberate practice in the acqui-sition of expert performance. Psychol. Rev. 100,363–406.

Ericsson, K. A., Roring, R. W., and Nandagopal,K. (2007). Giftedness and evidence for repro-ducibly superior performance: an accountbased on the expert performance frame-work. High Ability Stud. 18, 3–56. doi:10.1093/acprof:oso/9780199794003.003.0009

Gladwell, M. (2008). Outliers: The story of Success.New York, NY: Little, Brown, and Co.

Gobet, F., and Campitelli, G. (2007). The role ofdomain-specific practice, handedness and start-

ing age in chess. Dev. Psychol. 43, 159–172. doi:10.1037/0012-1649.43.1.159

Gobet, F., Campitelli, G., and Waters, A. J. (2002).Rise of human intelligence: comments on

Howard (1999). Intelligence 30, 303–311. doi:10.1016/S0160-2896(02)00083-1

Hambrick, D. Z., Oswald, F. L., Altmann, E. M.,Meinz, E. J., Gobet, F., and Campitelli, G.(2014). Deliberate practice: is that all it takes

to become an expert? Intelligence 45, 34–45. doi:

10.1016/j.intell.2013.04.001Howard, R. W. (1999). Preliminary real-world evi-

dence that average human intelligence really isrising. Intelligence 27, 235–250.

Salthouse, T. A. (2009). When does age-relatedcognitive decline begin? Neurobiol. Aging 30,507–514. doi: 10.1016/j.neurobiolaging.2008.09.023

Sujatha, S. (2013). Football keeps Magnus Carlsenawake. Available online at: http://archives.deccanchronicle.com/130821/sports-other-sports/article/football-keeps-magnus-carlsen-awake (Accessed29 June, 2014).

Syed, M. (2011). Bounce. London: Fourth Estate.

Conflict of Interest Statement: The authors declarethat the research was conducted in the absenceof any commercial or financial relationshipsthat could be construed as a potential conflict ofinterest.

Received: 29 June 2014; accepted: 23 July 2014;published online: 14 August 2014.Citation: Gobet F and Ereku MH (2014) Checkmate todeliberate practice: the case of Magnus Carlsen. Front.Psychol. 5:878. doi: 10.3389/fpsyg.2014.00878This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Gobet and Ereku. This is an open-access article distributed under the terms of the CreativeCommons Attribution License (CC BY). The use, dis-tribution or reproduction in other forums is permitted,provided the original author(s) or licensor are creditedand that the original publication in this journal is cited,in accordance with accepted academic practice. No use,distribution or reproduction is permitted which does notcomply with these terms.


http://www.forbes.com/sites/georgeanders/2014/01/17/chess-champ-carlsen-delights-silicon-valley-with-dry-wit/




http://archives.deccanchronicle.com/130821/sports-other-sports/article/football-keeps-magnus-carlsen-awake














doi: 10.3389/fpsyg.2014.00646

The influence of deliberate practice on musicalachievement: a meta-analysisFriedrich Platz1, Reinhard Kopiez2*, Andreas C. Lehmann3 and Anna Wolf 2

1 University of Music and Performing Arts, Stuttgart, Germany2 Hanover Music Lab, Hanover University of Music, Drama and Media, Hanover, Germany3 University of Music, Würzburg, Germany

Edited by:


Reviewed by:

Brooke Noel Macnamara, PrincetonUniversity, USAGary Edward McPherson, Universityof Melbourne, Australia

*Correspondence:

Reinhard Kopiez, Hanover Universityof Music, Drama and Media,Emmichplatz 1, 30175 Hanover,Germanye-mail: [email protected]

Deliberate practice (DP) is a task-specific structured training activity that plays a key role inunderstanding skill acquisition and explaining individual differences in expert performance.Relevant activities that qualify as DP have to be identified in every domain. For example,for training in classical music, solitary practice is a typical training activity during skillacquisition. To date, no meta-analysis on the quantifiable effect size of deliberate practiceon attained performance in music has been conducted. Yet the identification of aquantifiable effect size could be relevant for the current discussion on the role ofvarious factors on individual difference in musical achievement. Furthermore, a researchsynthesis might enable new computational approaches to musical development. Herewe present the first meta-analysis on the role of deliberate practice in the domain ofmusical performance. A final sample size of 13 studies (total N = 788) was carefullyextracted to satisfy the following criteria: reported durations of task-specific accumulatedpractice as predictor variables and objectively assessed musical achievement as the targetvariable. We identified an aggregated effect size of rc = 0.61; 95% CI [0.54, 0.67] for therelationship between task-relevant practice (which by definition includes DP) and musicalachievement. Our results corroborate the central role of long-term (deliberate) practice forexplaining expert performance in music.

Keywords: deliberate practice, music, sight-reading, meta-analysis, expert performance

INTRODUCTIONCurrent research on individual differences in the domain ofmusic is surrounded by controversial discussions: On the onehand, exceptional achievement is explained within the expert-performance framework with an emphasis on the role of struc-tured training as the key variable; on the other hand, researchersworking in the individual differences framework argue that (pos-sibly innate) abilities and other influential variables (e.g., workingmemory) may explain observable inter-individual differences (seeEricsson, 2014 for a detailed discussion). The expert-performanceapproach is represented by studies by Ericsson and cowork-ers (e.g., Ericsson et al., 1993) who assume that engaging inrelevant domain-related activities, especially deliberate practice(DP), is necessary and moderates attained level of performance.Deliberate practice is qualitatively different from work and playand “includes activities that have been specially designed toimprove the current level of performance” (p. 368). In a morecomprehensive and detailed definition, Ericsson and Lehmann(1999) refer to DP as a

“Structured activity, often designed by teachers or coaches withthe explicit goal of increasing an individual’s current level ofperformance. (· · · ) it requires the generation of specific goalsfor improvement and the monitoring of various aspects ofperformance. Furthermore, deliberate practice involves trying to

exceed one’s previous limit, which requires full concentration andeffort.” (p. 695)

In other words, we have to distinguish between mere experience(as a non-directed activity) and deliberate practice. An individ-ual’s involvement with a new domain entails the accumulationof experience, which may include practice components and leadto initially acceptable levels of performance. However, only theconscious use of strategies along with the desire to improve willresult in superior expert performance (Ericsson, 2006). Note thatin most studies DP is only indirectly estimated using durations oftask-relevant training activities that also include an unspecifiedproportion of non-deliberate practice components. The unre-flected use of the “accumulated deliberate practice” concept todenote durations of accumulated time spent in training activitiesis therefore misleading, because the measured durations mighttheoretically underestimate the true effect of deliberate practiceon attained performance. In the context of classical music per-formance, the task-relevant activity can often consist of sometype of solitary practice (e.g., studying repertoire or practicingscales) or the execution of a particular activity in a rehearsal ortraining context (e.g., sight-reading at the piano while coach-ing a soloist; receiving lessons). The theoretical framework forthe explanation of expert and exceptional achievement has beenvalidated in various domains and is widely accepted nowadays
















Platz et al. Deliberate practice in music performance

(Ericsson, 1996), as evidenced by the extremely high citation fre-quencies of key publications in this area. For example, accordingto Google Scholar, the study by Ericsson et al. (1993) has beencited more than 4000 times in the 20 years since its publication.As an internationally known proponent of research on giftedness,Ziegler (2009) concludes that even modern conceptions of gifted-ness research have integrated the perspective of expertise theory.However, controversial discussions persist (see Detterman, 2014).

In contrast, researchers relying more on talent-basedapproaches maintain that DP might not explain individualdifferences in performance sufficiently and emphasize innatevariables as the explanation for outstanding musical achievement,such as working memory capacity (Vandervert, 2009; Meinz andHambrick, 2010), handedness (Kopiez et al., 2006, 2010, 2012),sensorimotor speed (Kopiez and Lee, 2006, 2008), psychometricintelligence (Ullén et al., 2008), intrinsic motivation (Winner,1996), unique type of representations (Shavinina, 2009), orverbal memory (Brandler and Rammsayer, 2003). According toEricsson (2014), the predictive power of additional factors, suchas general cognitive abilities, is usually of small to medium sizeand diminishes as the level of expertise increases.

Although expertise theory provides convincing arguments forthe importance of structured training on expert skill acquisi-tion and achievement, no comprehensive quantification for theinfluence of DP on musical achievement has been presented sofar. A first and highly commendable attempt to estimate the“true” (population) effect of DP via estimates of durations ofaccumulated practice on musical achievement was published byHambrick et al. (2014) who identified a sample of eight studies fortheir review. However, their methodology, assumptions, and useof the term DP raise some issues that have to be resolved. Theseopen questions and concerns spawned our initial motivation forthe present meta-analysis.

REANALYSIS OF DATA PRESENTED IN Hambrick et al. (2014)First, we carefully studied the publication by Hambrick et al.(2014) (Table 1). Using Table 3 of their paper, we extractedthe correlations between training data and measures of musicperformance and entered these data into a meta-analysis soft-ware (Comprehensive Meta-Analysis, see Borenstein, 2010). Thisanalysis brought to light an aggregated efffect size value of r = 0.44for the influence of training data on musical performance (seeTable 1, for details). According to Cohen’s (1988) benchmarks,this corresponds to a large overall effect (see also Ellis, 2010, p. 41).Unlike Hambrick et al. (2014), we did not use the correlationvalues corrected for measurement error variance (attenuationcorrection) in the present paper because their correction of con-fidence intervals relied on the biased Fisher’s z transformation (seeHunter and Schmidt, 2004, Ch. 5) and not on the corrected sam-pling error variance for each individual correlation as suggested byHunter and Schmidt (2004, Ch. 3). Therefore, to allow for latercomparisons, we decided to use the uncorrected (attenuated)correlation as the basis for our analysis of heterogeneity.

The effect size, however, is not the only relevant parameterin a meta-analysis, and it should be examined in the light of apossible publication bias. To test for the strength of the result-ing effect size estimate, we conducted a test for heterogeneity for

Table 1 | Aggregation of data from Table 3 in Hambrick et al. (2014) for

the reanalysis of effect sizes regarding the influence of deliberate

practice on music performance.

Study N Variance r (95% CI) Relative

weight [%]

Lehmann and Ericsson, 1996 16 0.07 0.36 (−0.17, 0.73) 2.15

Meinz, 2000 107 0.01 0.41 (−0.24, 0.56) 17.22

Tuffiash, 2002 135 0.01 0.58 (−0.46, 0.68) 21.85

Kopiez and Lee, 2008 52 0.02 0.25 (−0.03, 0.49) 8.11

Ruthsatz et al., 2008—study 1 178 0.01 0.34 (−0.20, 0.46) 28.97

Ruthsatz et al., 2008—study 2A 64 0.01 0.31 (−0.07, 0.52) 10.10

Ruthsatz et al., 2008—study 2B 19 0.06 0.54 (−0.11, 0.80) 2.65

Meinz and Hambrick, 2010 57 0.02 0.67 (−0.50, 0.79) 8.94

MEAN AGGREGATED EFFECT SIZE

Fixed effect model 0.44 (−0.37, 0.50)

Random effects model 0.44 (−0.33, 0.55)

Aggregation of studies shows a large (I2 = 60.3%) and significant heterogeneity

(Q(7) = 17.7, p < 0.02).

the underlying sample of studies. Following Deeks et al. (2008),the I2 value describes the percentage of variance in effect sizeestimates that can be attributed to heterogeneity rather than tosampling error. The I2 value of 60.3 obtained for the Hambricket al. (2014) sample of studies implied that it “may representsubstantial heterogeneity” (Deeks et al., 2008, p. 278). The mainreason for possible heterogeneity, in our opinion, could be aless selective inclusion with resulting inconsistent predictor andtarget variables. For example, in their study on the acquistionof expertise in musicians, Ruthsatz et al. (2008) used inconsis-tent (non-standardized) indicators for the estimation of musicalachievement that made it difficult to compare the observed dif-ferences in performance: In Study 1, the band director’s auditionscores for each of the high school band members were ranked andused as individual indicators of musical achievement; in Study2A, audition scores from the admission exam were used as theoutcome variable; and in Study 2B, a music faculty memberrated the students’ general musical achievement. In no instancewas a standardized performance task used as the target vari-able. Unfortunately, no information was reported on the ratingreliabilities.

Although our reanalysis of Hambrick et al.’s (2014) review con-firmed a large effect size for the relation between training dataand musical achievement, this finding still underestimates the“true” value. In order to arrive at a convincing effect size fordeliberate practice in the domain of music we also aggregatedstudies, but invested great effort in the selection of studies for ourmeta-analysis. As will be shown below, our meta-analysis was notaffected by potential publication bias and heterogeneity. We alsoapplied transparent and consistent criteria for study selection asthis is one of the most important prerequisites for the aggregationof studies.

CHOICE OF METHODTwo methods are available to evaluate past research: (a) a narra-tive and systematic review and (b) a meta-analysis. The narrativereviewer uses published studies, reports other authors’ results in






his or her own words and draws conclusions (Ellis, 2010, p. 89).A systematic review is also sometimes referred to as a “quali-tative review” or “thematic synthesis” (Booth et al., 2012) andnecessitates a comprehensive search of the literature. The disad-vantage of this approach is that it depends on the availabilityof results published in established journals and tends to showa publication bias toward the Type I error (false positive). Thereason for this is that journals prefer to publish studies with sig-nificant results, and negative findings or null results have a lowerprobability of publication (Masicampo and Lalande, 2012). Inthe field of music, narrative reviews on the influence of DP onmusical achievement play an important role and have been con-ducted in the last two decades (Lehmann, 1997, 2005; Howe et al.,1998; Sloboda, 2000; Krampe and Charness, 2006; Lehmann andGruber, 2006; Gruber and Lehmann, 2008; Campitelli and Gobet,2011; Hambrick and Meinz, 2011; Nandagopal and Ericsson,2012; Ericsson, 2014).

The other approach is that of a meta-analysis. Here, studiesare included following “pre-specified eligibility criteria in orderto answer a specific research question” (Higgins and Green, 2008,p. 6). Within the meta-analytic approach, studies’ effect sizes haveto be weighted before they are aggregated. Every study’s effect sizeweight then reflects its degree of precision as a function of sam-ple size (Ellis, 2010). Consequently, studies with smaller samplesizes, particularly in combination with larger variation, will resultin smaller weights compared to studies with larger sample sizesand more narrow variation. These weights of the individual stud-ies then function as estimators of precision. If these weights differmarkedly from each other, statistical heterogeneity is present. Thefinal result of a meta-analysis is the weighted mean effect sizeacross all studies included. Compared to an individual study’seffect size, this weighted mean effect size represents a more pre-cise point estimate as well as an interval estimate surroundingthe effect size in the population (Ellis, 2010, p. 95). Moreover,a meta-analysis generally increases statistical power by reducingthe standard error of the weighted average effect size (Cohn andBecker, 2003). Researchers who use meta-analysis techniques havetwo goals: First, they want to arrive at an interval of effect sizeestimation in a population based on aggregated effect sizes ofindividual studies; second, they want to give an evidence-basedanswer to those questions that reviews or replication studies can-not give in part due to their arbitrary collection of significant andinsignificant results.

Despite the fact that meta-analyses have been shown to be animportant constituent for the production of “verified knowledge”(Kopiez, 2012), they have only recently been applied to varioustopics in music psychology (e.g., Chabris, 1999; Hetland, 2000;Pietschnig et al., 2010; Kämpfe et al., 2011; Platz and Kopiez, 2012;Mishra, 2014). To date, there has been no formal meta-analysisconcerning the influence of DP on attained music performance.

GOAL OF THE PRESENT STUDYThe aim of our study was two-fold: First, by means of a systematicliterature review we wanted to identify all relevant publicationsthat might help us answer the question of how strongly task-specific practice influences attained music performance. Second,we wanted to quantify the effect of DP on music performance in

terms of an objectively computed effect size. This effect size is animportant component for the development of a comprehensivemodel for the explanation of individual differences in the domainof music. Although this meta-analysis is supposed to reveal the“true” effect size of deliberate practice on musical achievement,for theoretical reasons it is possible that it is still underestimatingthe upper bound of deliberate practice (see Future Perspectives).

MATERIALS AND METHODSThe study was conducted in three steps: First, to arrive at arelevant sample of selected studies, we conducted a systematicreview (Cooper et al., 2009) that helped to control for publica-tion bias (Rothstein et al., 2005). In the second step, we identifiedeach study’s predictor and outcome variable in line with Ericsson(2014), and we identified all artifactual confounds that mightattenuate the studies’ outcome measures (Hunter and Schmidt,2004, p. 35). Third, we carried out a meta-analysis of individuallycorrected (disattenuated) correlations as well as a quantifica-tion of its variance (Hunter and Schmidt, 2004; Schmidt andLe, 2005) to obtain the true mean score correlation (ρ) betweenmusic-related practice and musical achievement.

SAMPLE OF SELECTED STUDIESOur sample of selected studies for the subsequent meta-analysiswas the outcome of a systematic literature search which had led toa preliminary corpus of selected studies (see Figure 1A). Due to awide variety of methodological approaches, and for the purposeof later generalizability of our meta-analytical results, we decidedto select only studies with comparable experimental designs.Therefore, in the next step of generating a sample, we excludedall studies from the preliminary corpus that did not meet all ofour selection criteria (see Figure 1B). Consequently, our prelim-inary corpus of n = 102 studies dwindled to the final sample ofn = 13 studies which served as input for the meta-analysis.

LITERATURE SEARCHThe acquisition of studies for our systematic review derived from(a) the search for relevant databases of scientific literature, (b)queries of conference proceedings, and (c) personal communi-cations with experts in the field of music education or musicaldevelopment. First, a database backward and forward search forliterature was conducted in January 2014 (Figure 1A). To controlfor publication bias (see Rothstein et al., 2005), we considered alarge variety of databases for our literature search: peer-reviewedstudies in the field of medical and neuroscientific (PubMed),psychological (PsycINFO), educational (ERIC), social (ISI), andmusicological research (RILM). To avoid an overestimation of theeffect size due to possibly unpublished results (Rosenthal, 1979),so-called “gray literature” (Rothstein and Hopewell, 2009) withoften non-significant study results, we also searched doctoral dis-sertations (DAI), proceedings or newspaper articles (PsycEXTRA)as well as book chapters containing psychological study results(PsycBOOKS).

Studies were excluded from the preliminary corpus if they didnot conform with at least one of the following three descrip-tors (Figure 1A): (1) “music” AND “deliberate practice,” (2)“music” AND “formal practice,” (3) “music” AND “expertise.”





FIGURE 1 | Arriving at a study sample for the meta-analysis. In the firststep (A), a search for literature was based on selected descriptors applied toeight data bases. This resulted in a preliminary corpus of 102 studies. In the

second step (B), studies were evaluated and selected for meta-analysisaccording to seven criteria. N = 13 studies matched all criteria and wereincluded into the meta-analysis.

In addition, we included in the preliminary corpus those music-related studies which cited Ericsson et al.’s (1993) first extensivereview of skill acquisition research. Finally, authors who had con-ducted experimental studies on predictors of music achievementwere contacted and queried for currently unpublished correla-tional data involving music-related deliberate practice and musi-cal achievement. In total, our initial literature search resulted in apreliminary corpus of 102 studies (Figure 1A).

CRITERIA-RELATED LITERATURE SELECTIONWhile Hambrick et al. (2014) performed a more intuitive search,resulting in a significant heterogeneity of the study sample, theaim of our method was to arrive at a homogenous sample of

pertinent studies. To this end, we selected studies based on objec-tive criteria which we derived from the theoretical frameworkof expert performance according to Ericsson et al. (1993). Thus,studies were successively removed from the preliminary corpus ofstudies if they did not meet all the criteria shown in Figure 1B. Asa result of our study selection (see Table 2), we identified studieswhich met the following 6 criteria: (1) they followed a hypothesis-testing design; (2) they contained a correlation between accumu-lated deliberate practice and a corresponding task-related level ofmusical achievement; (3) the amount of relevant practice had tobe accrued across at least 1 year, (4) musical performance had tobe measured by means of objective criteria such as a computer-based assessment (e.g., scale analysis by Jabusch et al., 2004) or






expert evaluation based on psychometric scales (e.g., Hallam,1998). (5) Furthermore, studies were excluded if they did notcontain sufficient statistical information for effect size calculationor estimation. (6) Finally, in the case of duplicate publication ofdata (as happens when original articles are also published in chap-ter form), study results were considered only once for effect sizeaggregation in the meta-analysis.

Following our selection criteria n = 89 studies had to beexcluded from our preliminary corpus. Our final sample size wasthus n = 13 studies, comprising results from peer-reviewed stud-ies as well as “gray” literature from 1992 to 2012 (see Table 2).For comparison, Hambrick et al.’s (2014) sample size of studiesincluded in his review was n = 8.

PROCEDUREAccording to Hunter and Schmidt (2004, p. 33), the aim of apsychometric meta-analysis is two-fold: namely, to uncover the

variance of observed effect sizes (s2r )—in our study, this was the

variance of observed correlations between the task-related prac-tice (predictor) and musical achievement (outcome variable)—and to estimate the supposedly “true” effect size distributionin the population

(σ2

ρ

). The use of the term “psychometric”

refers to the idea in classical testing theory (Gulliksen, 1950) thatevery observed correlation is subject to an attenuation due to theimperfect measurement of variables, sampling error, and furtherartifacts (for an overview see Hunter and Schmidt, 2004, p. 35). Ifthe influence of all such artifactual influences on an observed cor-relation are known (ro), each study’s correlation can be correctedfirst for its individual attenuation bias (rc). In a subsequent step,the population variance of the “true” correlation (σ2

ρ) is estimatedby subtracting the observed variance of corrected correlations(s2

rc) from the observed variance attributable to all attenuating

factors (s2ec

). In the case of a perfect concordance between the

observed variance of corrected correlations (s2rc

) and the observed

Table 2 | Studies, included in meta-analysis.

ID Study Comments

Kornicke, 1992 Kornicke, L. E. (1992). An exploratory study of individual difference variables in pianosight-reading achievement (Doctoral Dissertation, Indiana University, Ann Arbor, USA).Available from ProQuest Dissertations and Theses database. (UMI No. 9301458).

Ericsson et al., 1993—Study II Ericsson, K. A., Krampe, R. T., and Tesch-Römer, C. (1993). The role of deliberatepractice in the acquisition of expert performance. Psychological Review 100, 363–406.

Two studies reported; onlydata of study II wasconsidered.

Lehmann and Ericsson, 1996 Lehmann, A. C., and Ericsson, K. A. (1996). Performance without preparation: structureand acquisition of expert sight-reading and accompanying performance.Psychomusicology 15, 1–29.

Krampe and Ericsson,1996—Study I

Krampe, R. T., and Ericsson, K. A. (1996). Maintaining excellence: deliberate practiceand elite performance in young and older pianists. Journal of Experimental Psychology:General 125, 331–359.

Two studies reported; onlydata of study I wasconsidered.

Hallam, 1998 Hallam, S. (1998). The predictors of achievement and dropout in instrumental tuition.Psychology of Music 26, 116–132.

Meinz, 2000 Meinz, E. J. (2000). Experience-based attenuation of age-related differences in musiccognition tasks. Psychology and Aging 15, 297–312.

Tuffiash, 2002 Tuffiash, M. (2002). Predicting individual differences in piano sight-reading skill:practice, performance, and instruction. Unpublished master’s thesis, Florida StateUniversity, Tallahassee, FL.

McPherson, 2005 McPherson, G. E. (2005). From child to musician: skill development during thebeginning stages of learning an instrument. Psychology of Music 33, 5–35.

Author contacted for data.

Jabusch et al., 2007 Jabusch, H.-C., Yong, R., and Altenmüller, E. (22–23 Nov. 2007). Biographical predictorsof music-related motor skills in children pianists. Paper presented at the InternationalSymposium on Performance Science, Porto.

Kopiez and Lee, 2008 Kopiez, R., and Lee, J. I. (2008). Towards a general model of skills involved in sightreading music. Music Education Research 10, 41–62.

Jabusch et al., 2009 Jabusch, H. C., Alpers, H., Kopiez, R., Vauth, H., and Altenmüller, E. (2009). Theinfluence of practice on the development of motor skills in pianists: a longitudinal studyin a selected motor task. Human Movement Science 28, 74–84.

Meinz and Hambrick, 2010 Meinz, E. J., and Hambrick, D. Z. (2010). Deliberate practice is necessary but notsufficient to explain individual differences in piano sight-reading skill: the role ofworking memory capacity. Psychological Science 21, 914–919.

Kopiez et al., 2012—Study II Kopiez, R., Jabusch, H.-C., Galley, N., Homann, J.-C., Lehmann, A. C., and Altenmüller,E. (2012). No disadvantage for left-handed musicians: the relationship betweenhandedness, perceived constraints and performance-related skills in string players andpianists. Psychology of Music 40, 357–384.

Two studies reported; onlydata of study II wasconsidered.





variance attributable to all artifacts (s2ec

), there is no population

variance left to be explained (σ2ρ = 0). Then all studies’ effect

sizes in the meta-analysis are homogenous and assumed to derivefrom one single population effect (Hunter and Schmidt, 2004, p.202). Therefore, we will first identify each study’s theoreticallyappropriate predictor and outcome variable as well as reliabilityinformation for both variables in order to calculate effect size andestimate artifactual influence.

IDENTIFICATION OF PREDICTORS AND OUTCOME VARIABLESAlthough accumulated deliberate practice on an instrument hasbeen identified as a generally important biographical predictorin the acquisition of expert performance (Ericsson et al., 1993),it is sometimes erroneously considered a catch-all predictor forachievement in music-specific tasks. However, as Ericsson clearlystates, “it is not the total number of hours of practice that mat-ter, but a particular type of practice [emphasis by the third author,AL] that predicts the difference between elite and sub-elite ath-letes” (Ericsson, 2014, p. 94). For example, according to Lehmannand Ericsson (1996) as well as Kopiez and Lee (2006, 2008),sight-reading performance as a domain-specific task of musicalachievement should be less well predicted by accumulated genericdeliberate practice in piano playing (i.e., solitary practice) thanby the accumulated amount of task-specific deliberate practicein the field of accompanying and sight-reading. Therefore—and in contrast to Hambrick et al.’s (2014) procedure—for eachstudy we identified the most corresponding predictor variable.For example, the researcher might have summed up the num-ber of pieces sight-read (Kornicke, 1992, p. 133), determinedthe size of the accompanying repertoire (Lehmann and Ericsson,1996, p. 29), counted the number of accompanying performances(Meinz, 2000, p. 301), reported cumulated piano accompanyingperformances (Tuffiash, 2002, p. 81), calculated the accumulatedsight-reading expertise until the age of 18 (Kopiez and Lee, 2008,p. 49) or aggregated the durations of accompaniment and hoursof specific sight-reading practice (Meinz and Hambrick, 2010, p.3). Information on the task-specific accumulated practice dura-tion until the age of 18 or 20 years was used in the case of Ericssonet al. (1993, p. 386), Krampe and Ericsson (1996, p. 347), andKopiez and Lee (2008, p. 49). In the absence of such data, we usedthe total accumulated practice time (at the time of the data collec-tion) instead (e.g., in the case of Hallam, 1998, p. 124; McPherson,2005, author contacted for data; Jabusch et al., 2007, p. 366; andKopiez et al., 2012, p. 372).

In addition to the predictor variable, the measurement ofthe outcome variable should be representative of the investi-gated skill (Ericsson, 2014). Consequently, inter-onset evenness inscale-playing as well as performed (rehearsed) music were iden-tified as truly domain-specific tasks of musical achievement inour sample of studies on music performance. Here, participants’performances were measured either by a reliable psychologi-cal evaluation based on psychometric scale construction (e.g.,Kornicke, 1992) or by an objective, computer-based, physicalmeasurement such as obtaining the number of correctly per-formed notes (e.g., Lehmann and Ericsson, 1996) or identifyingthe inter-onset evenness of scale-playing (e.g., Ericsson et al.,1993; Krampe and Ericsson, 1996; Jabusch et al., 2007). In the

case of multiple tasks, as was the case in Ericsson et al. (1993, p.386) as well as in Krampe and Ericsson (1996, p. 347), we decidedto choose the task with the stronger measurement reliability, thehighest difficulty and the highest discrimination ability for musi-cal achievement (different movements with each hand (Ericssonet al., 1993, p. 386), simultaneously [Exp. 1], see Krampe andEricsson, 1996).

RELIABILITY OF IDENTIFIED PREDICTORS AND OUTCOME VARIABLESFor the purpose of adjusting the correlation coefficient of theobserved studies for attenuation, the measurement error in thepredictor as well as in the outcome variable had to be identified(Hunter and Schmidt, 2004, p. 41). As shown in Table 3, only asmall number of studies reported information on the reliabilityfor either the predictor or the outcome variable. Specifically, onlyTuffiash (2002, p. 36) reported test-retest reliability in cumulativepiano accompaniment performance (rxx = 0.91) for the quan-tification of measurement error in the predictor variable. Histest-retest reliability estimations were similar to those reportedin Bengtsson et al. (2005, p. 1148), who stated a mean test-retestreliability rxx = 0.89 for the estimation of accumulated deliber-ate practice obtained from retrospective interviews. Thus, whenno reliability was reported for the predictor variable, we used themean correlation of test-retest reliability according to Bengtssonet al. (2005) to estimate the imperfection of the predictorvariable.

To quantify measurement error in the outcome variable, weused the Cronbach’s alpha reported in Kornicke (1992, p. 109)for the inter-rater reliability of the sight-reading test and inMcPherson (2005, p. 13) for performing rehearsed music. InKrampe and Ericsson (1996, p. 339) and Meinz and Hambrick(2010, p. 4), Cronbach’s alpha of the construct reliability for thepsychometric measurements could be copied from the respectivepapers. Finally, in the case of Tuffiash (2002, p. 28) we computeda mean correlation on the basis of all the test-retest reliabilities ofsight-reading tests the author reported. For studies in which nomeasurement error was stated for the outcome variable, we esti-mated the reliability of the outcome variable’s measurement: Toestimate the reliability of experts’ performance ratings for the out-come variable in Lehmann and Ericsson (1996) and Kopiez andLee (2008), we used the intercorrelations between the expert judg-ment of overall impression and the amount of correctly playednotes (ryy = 0.88) as reported in Lehmann and Ericsson (1993, p.190). In the cases of Ericsson et al. (1993), Jabusch et al. (2007,2009) and Kopiez et al. (2012), we estimated ryy = 0.91 as theconstruct reliability according to Spector et al. (in revision); theycomputed a mean correlation of test-retest reliability for Jabuschet al.’s (2004) measurement of note-evenness in scale playing. Thesame test-retest reliability of the scale-analysis by Spector et al.(in revision) was used for the estimation of the test-retest relia-bility for the ABRSM in Hallam (1998). Along the lines of Bergee(2003), we underestimated the disattenuated correlation by usingryy = 0.91 and obtained a more conservative correction. Finally,a reliability estimate of ryy = 0.96 for Meinz (2000) was commu-nicated by the author and also reported in Hambrick et al. (2014,p. 6). In summary, all studies showed a weak attenuation with a1–17% downwards bias (see Table 4, column A).






Table 3 | Reported effect size data on the relationship between indicators of deliberate practice and objective measurement of musical

achievement.

ID Study design Effect size data

Sample Predictor Performance measure Sig. report Reliability*

n r p rxx ryy

Kornicke, 1992 College level pianists Composite number ofpieces sight-read

Expert rating ofsight-reading performance

73 0.50 0.99

Ericsson et al.,1993—study II

University musicmajors (pianists)

Accumulated practice Evenness of inter-onsetintervals

24 −0.857 <0.01

Lehmann andEricsson, 1996

University musicstudents

Accompanying score Number of correctlyperformed notes

16 0.72 <0.01

Krampe and Ericsson,1996—study I+

Beginning toprofessional pianists

Accumulated practice(until age of 20)

Evenness of inter-onsetintervals

48 −0.62 <0.01 0.97

Hallam, 1998 Beginners Accumulated practicetime

Associated board of theroyal schools music(ABRSM)

109 0.67 <0.01

Meinz, 2000 Beginning to advancedpianists

Number ofaccompanyingperformances


107 0.57 <0.01

Tuffiash, 2002 Undergraduate musicand non-music majors

Cumulative pianoaccompanimentperformances

Expert ratings of musicperformances

135 0.426 <0.01 0.91 0.75

McPherson, 2005 Beginners Accumulated practicetime (over 3 years)

Expert rating of performedrehearsed music

99 0.568 <0.01 0.92

Jabusch et al., 2007+ School-aged children Accumulated practicetime


30 −0.46 <0.05

Kopiez and Lee, 2008 Piano major studentsand graduates

Accumulatedsight-reading expertise(until age of 18)

Sight-reading achievement 52 0.359 <0.01

Jabusch et al.,2009+◦◦

University musicstudents

Life-time deliberatepractice


19 −0.44 <0.01

Meinz and Hambrick,2010◦◦◦

Beginners to advancedpianists

Accumulatedaccompaniments andhours of deliberatesight-reading practice


57 0.56 <0.01 0.99

Kopiez et al.,2012—Study II+◦

University musicstudents (piano major)

Accumulated practicetime


19 −0.42 <0.05

+Absolute values were used in meta-analysis.◦Aggregated correlation based on all four correlations between accumulated deliberate practice and outcome variable.◦◦Aggregated correlation based on two reported correlations between accumulated life-time deliberate practice and outcome variable.◦◦◦According to Lehmann and Ericsson (1996) the mean correlation of accompaniments (r = 0.63) and hours of deliberate sight-reading practice (r = 0.48) was used

as task-specific predictor for sight-reading performance.*Reliability coefficients reported in studies; assumed reliability (if not reported) of predictor variable used for attenuation correction in meta-analysis: rxx = 0.89;

assumed reliability (if not reported) of outcome variable (ryy ) for attenuation correction in meta-analysis: Ericsson et al., 1993 (ryy = 0.91), Lehmann and Ericsson,

1996 (ryy = 0.88), Hallam, 1998 (ryy = 0.91), Meinz, 2000 (ryy = 0.96), Jabusch et al., 2007 (ryy = 0.91), Kopiez and Lee, 2008 (ryy = 0.88), Jabusch et al., 2009

(ryy = 0.91), Kopiez et al., 2012 (ryy = 0.91).

STATISTICAL REANALYSIS AND META-ANALYSIS WITHCORRELATIONS CORRECTED FOR ARTIFACTSAll studies reported correlations that could be used for quantify-ing the effect of deliberate practice on the musical achievement

(see Table 3). Meinz and Hambrick (2010) reported multi-ple predictors of sight-reading skill along the theoretical out-line for the acquisition of sight-reading skill (Lehmann andEricsson, 1996; Kopiez and Lee, 2006). We aggregated the two





predictors, number of accompanying events/activities (r = 0.63)and hours of sight-reading practice (r = 0.48), into a meancorrelation (r = 0.56) to be used as a global predictor forsight-reading performance (see Table 3). As a result of a 2 ×2 experimental design, four correlations of pianists’ accumu-lated task-specific practice times and scale performances werereported in Kopiez et al. (2012). Again, the four individual cor-relations

(rLi = −0.47; rLo = −0.23; rRi = −0.46; rRo = −0.50

)

were aggregated to the study’s effect size (r = −0.42) (Kopiezet al., 2012, Table 6 on p. 372; see comment on negative valuesbelow). Finally, in the case of Jabusch et al. (2009, p. 77), twocorrelations between total life-time practice and music perfor-mance (as measured by evenness in scale playing on various dateswith a distance of 1 year; r1 = −0.47; r2 = −0.40) were reported.We calculated and used the mean correlation (|r| = 0.44) in ourmeta-analysis.

Jabusch et al.’s (2004) scale-playing paradigm generallyresulted in negative correlations (see Table 3). Since the authorsreport the median of the scale-related inter-onset interval stan-dard deviation (medSDIOI) as an indicator for evenness, a lowmedSDIOI signals high evenness. A positive association betweenaccumulated practice times and the medSDIOI can still be pos-tulated: the longer the pianist’s deliberate practice durations,the smaller the degree of unevennes. For the sake of simplicitywe used the absolute values of the correlations reported in ourmeta-analysis (this also applies to Ericsson et al., 1993; Krampeand Ericsson, 1996; Jabusch et al., 2007, 2009; Kopiez et al.,2012).

Finally, the observed correlations as well as the reliabilities ofpredictor and outcome variables were entered into the Hunter-Schmidt Meta-Analysis software (Schmidt and Le, 2005) so thatwe could correct all observable correlations for artifacts (Hunterand Schmidt, 2004, p. 75) within the meta-analysis and estimatethe population correlation for the “true” effect size (see Table 4).

RESULTSSTATISTICAL PROCEDUREThe observed correlation (ro) for each study was transformed intoits disattenuated rc value. This disattenuation procedure is basedon the assumption that the observed correlation (ro) comprisesthe “true” value plus the influence of a measurement error thatdepends on the reliability of both the predictor (rxx) and out-come (ryy) variable. According to Hunter and Schmidt (2004),the ro value has to be corrected for limited reliability of both vari-ables, and this correction is implemented in the Hunter-SchmidtMeta-Analysis Programs (see Schmidt and Le, 2005). Detailedresults with all steps and for each study are shown in Table 4.It is remarkable that 81.2% of the complete variance in all cor-rected correlations was attributable to the artifacts, a findingwhich leaves no residual variance to be explained (for an expla-nation, see Hunter and Schmidt, 2004, p. 401). In other words,our meta-analysis is based on an homogenous corpus of data(Q(12) = 8.19, p = 0.77; I2 = 0.00%) which is the outcome ofa careful sampling and study selection, guided by the criteriaof task-specific practice and objective measurements of musicperformance.

Table 4 | Statistical values of the meta-analysis.

ID N ro Reliability A Var(eo) Var(ec) w Weight [%] rc

rxx ryy

Kornicke, 1992 73 0.50 0.89 0.99 0.94 0.01 0.01 64.32 10.10 0.53Ericsson et al., 1993—study II 24 0.86 0.89 0.91 0.90 0.02 0.03 19.44 3.05 0.96Lehmann and Ericsson, 1996 16 0.72 0.89 0.88 0.88 0.03 0.04 12.53 1.97 0.81Krampe and Ericsson, 1996—study I 48 0.62 0.89 0.97 0.93 0.01 0.01 41.44 6.51 0.67Hallam, 1998 109 0.67 0.89 0.91 0.90 0.00 0.01 88.28 13.87 0.74Meinz, 2000 107 0.57 0.89 0.96 0.92 0.00 0.01 91.42 14.36 0.62Tuffiash, 2002 135 0.43 0.91 0.75 0.83 0.00 0.01 92.14 14.47 0.52McPherson, 2005 99 0.57 0.89 0.92 0.90 0.01 0.01 81.06 12.73 0.63Jabusch et al., 2007 30 0.46 0.89 0.91 0.90 0.02 0.02 24.30 3.82 0.51Kopiez and Lee, 2008 52 0.36 0.89 0.88 0.88 0.01 0.01 40.73 6.40 0.41Jabusch et al., 2009 19 0.44 0.89 0.91 0.90 0.03 0.03 15.39 2.42 0.49Meinz and Hambrick, 2010 57 0.56 0.89 0.99 0.94 0.01 0.01 50.22 7.89 0.60Kopiez et al., 2012—study II 19 0.42 0.89 0.91 0.90 0.03 0.03 15.39 2.42 0.47

N, sample size; ro, observed correlation (Hunter and Schmidt, 2004, p. 96); rxx , reliability of predictor variable (error of measurement in the predictor variable, Hunter

and Schmidt, 2004, p. 96); ryy , reliability of outcome variable (error of measurement in the outcome variable, Hunter and Schmidt, 2004, p. 96); A, attenuation

factor (ro/rc , Hunter and Schmidt, 2004, p. 118); Var(eo), sampling error variance of each study’s uncorrected correlation (Hunter and Schmidt, 2004, p. 87); Var(ec),

sampling error variance of each study’s corrected correlation (Hunter and Schmidt, 2004, p. 119); w, study weight (Hunter and Schmidt, 2004, p. 125); rc , corrected

study correlation (Hunter and Schmidt, 2004, p. 118); weighted mean observed correlation ro = 0.54 (Hunter and Schmidt, 2004, p. 81); frequency-weighted average

squared error S2r = 0.01 (Hunter and Schmidt, 2004, p. 81); mean true score correlation ρ = 0.61 (Hunter and Schmidt, 2004, p. 125); variance of true score

correlations S2ρ = 0 (Hunter and Schmidt, 2004, p. 126); observed variance of the corrected correlations S2

rc = 0.01 (Hunter and Schmidt, 2004, p. 126); variance in

corrected correlations attributable to all artifacts S2ec

= 0.01 (Hunter and Schmidt, 2004, p. 126); complete variance in corrected correlations (81.2%) is attributable

to all artifacts (Hunter and Schmidt, 2004, p. 401); Q-test on study homogeneity as well as I2 suggest no significant variation across studies (I2 = 0.00; Q(12) =8.19, p = 0.77).






MAIN OUTCOMEThe result from 13 studies regarding the effect of the indica-tors of DP on musical achievement is summarized in Figure 2using a forest plot. Our meta-analysis yielded an average aggre-gated corrected effect size of rc = 0.61, with CI 95% [0.54, 0.67].According to Cohen’s benchmarks (1988, p. 80), this correspondsto a large effect. The size of the squares in the forest plot indi-cates each study’s weight and error bars delimit the 95% CI. Theremarkably strong relationship between task-specific practice andmusical achievement as measured by objective means is only onefacet of the aggregated and corrected correlations. Another facetof the results is the 95% CI as a measure of dispersion for thepopulation effect which is rather narrow [0.54, 0.67] and positive.This feature indicates the stability of our finding. The forest plotalso shows that the aggregated correlation is not biased by one ortwo studies with extreme relative weights. Rather, a total of 4 stud-ies (Hallam, 1998; Meinz, 2000; Tuffiash, 2002; McPherson, 2005)with high relative weights contribute 50% to the aggregated result.

TEST FOR PUBLICATION BIASEvidence suggests that due to their selective decision processesand preference for significant results, peer-reviewed journals onlypartially reflect research activities (Rothstein et al., 2005). Thisso-called publication or availability bias is an indicator for theexistence of unpublished results, and it is a sign of how stronglythose unpublished studies could influence the results of a meta-analysis. To detect the presence of a systematic selection bias ofpublications, we used the so-called funnel plot (Egger et al., 1997)(see Figure 3). If publication bias is present, the distribution ofresults will form an asymmetrically shaped funnel. Fortunately,Figure 3 shows a nearly symmetrical distribution of effect sizes inrelation to the standard error (the indicator of precision). Withthe exception of one, the effect sizes lie within the funnel’s shapeand are centered symmetrically around the aggregated mean ofrc = 0.61. Such considerably low bias is one of the strengths ofour meta-analysis and the result of carefully defined criteria forinclusion (see Figure 1).

DISCUSSIONOne of the main results of our meta-analysis is the identificationof a reliable, aggregated correlation between task-relevant prac-tice and objectively measured musical achievement. Although thecentral parameter of our analysis of 13 studies is similar to theone calculated by Hambrick et al. (2014) on the basis of 8 stud-ies, there are some marked differences between both approaches.Our results may currently represent the best estimate of thiscorrelation given the published data and methodological tools.

COMPARISON OF OUR FINDINGS TO THOSE BY Hambrick et al. (2014)An important step in the use of correlation coefficients in meta-analyses is the correction for attenuation (Hunter and Schmidt,

FIGURE 3 | Funnel plot of studies’ effect sizes (rc ) against standard

error of effect sizes as a test for publication bias.

FIGURE 2 | Forest plot of corrected effect sizes for individual studies and of the aggregated mean effect size (rc = 0.61, 95% CI [0.54, 0.67]) based on

the total number of N = 788 participants. Error bars indicate 95% CI; the size of the squares corresponds to the relative weight of the study.





2004). It considers the reliability of the outcome and predictorvariables in a study. Although we chose conservative estimatesof reliability for the disattenuation procedure in the presentpaper, our resulting correlation value is higher (rc = 0.61) thanHambrick et al.’s (2014) (rc = 0.52), and it covers a smaller con-fidence interval (95% CI [0.54, 0.67]) compared to theirs (95%CI [0.43, 0.64]). Therefore, we conclude that our meta-analysis isa more reliable approximation of the “true” correlation betweentask-relevant practice (including DP) and musical achievement.

In some instances, the predictors we used were different fromthose Hambrick et al. (2014) had used for their study. Forexample, they selected the value of ro = 0.25 from the sight-reading study by Kopiez and Lee (2008). However, this correlationbetween task-relevant study (i.e., sight-reading expertise) andactual sight-reading achievement was based on the lifetime accu-mulated practice time in sight-reading (up to the time of datacollection). In line with the criteria for the calculation of accu-mulated practice time employed in Ericsson et al. (1993); Ericssonet al. (Study II, see Table 3), and for reasons of comparability, weused the correlation between accumulated sight-reading exper-tise up to the age of 18 years and sight-reading performance (ro =0.36; Kopiez and Lee, 2008) for our meta-analysis. Life-time accu-mulated practice durations were only used when no informationon the task-specific accumulated practice time until the age of 18or 20 years could be obtained from the studies. We believe thatthe careful selection of studies and variables based on selectioncriteria of objective measurement for the outcome (performance)variable and clear calculations of accumulated practice durationsare the main reasons for the differences between Hambrick et al.’sresults and ours.

THE ROLE OF POSSIBLE FURTHER MODERATING VARIABLES ONPERFORMANCEThe discussion on the influence of variables other than studydurations that might influence musical achievement is ongoingand interesting. Here, we wish to comment on the tendency ofauthors to use headings for publications that can be misleadingfor the uninformed reader. For example, Meinz and Hambrick(2010) insinuate that there might be (heritable) variables whichhave a significant influence on musical achievement, and theysuggest working memory capacity as such an influential factor.Yet, their main finding regarding the central role of various formsof relevant practice on sight-reading achievement (within a rangefrom ro = 0.37 to 0.67) implies that working memory capacitycan only contribute a smaller proportion of the variance (ro =0.28). Although the authors conclude “that deliberate practiceaccounted for nearly half of the total variance in piano sight-reading performance” (Meinz and Hambrick, 2010, p. 914), thearticle title, “Limits on the Predictive Power of Domain-SpecificExperience and Knowledge in Skilled Performance,” defames therole of deliberate practice. A second case is the publication byRuthsatz et al. (2008) in which the authors found a low corre-lation between general intelligence (IQ) and musical achievementof ro = 0.25 (Study 1), 0.11 (Study 2A), and −0.01 (Study 2B)but a large one between accumulated practice time and musi-cal achievement (ro = 0.34 [Study 1], 0.31 [Study 2A], and 0.54[Study 2B]). Their combination of “other” variables exceeds the

influence of deliberate practice times only when the aggregatedcorrelations of IQ and music audiation are compared with theinfluence of the individual predictor of practice. However, it iswell-known that Gordon’s tests of audiation (AMMA), whichRuthsatz uses, is influenced by musical experience and thusalready captures effects of DP. In light of such findings, theauthors’ claim that “higher-level musicians report significantlyhigher mean levels of characteristics such as general intelligenceand music audiation, in addition to higher levels of accumulatedpractice time” (Ruthsatz et al., 2008, p. 330) is grossly misleading.

Another argument for a differentiated view of our findingsarises from the erroneous interpretation of r (or rc) values as r2

values known from common variance. For example, Hambricket al. (2014, p. 7) state: “On average across studies, deliberatepractice explained about 30% of the reliable variance in musicperformance.” However, according to Hunter and Schmidt (2004,p. 190), this is a problematic interpretation with regard to findingsfrom a meta-analysis, because the r2 value is “related only in a verynonlinear way to the magnitudes of effect sizes that determinetheir impact in the real world.” Instead, relationships betweenvariables should be interpreted in terms of linear relationships.Therefore, we could illustrate the relevance of our meta-analyticalfinding by means of a correlation simulation based on a samplesize of N = 788 and a given correlation of rc = 0.61. Figure 4displays this simulation with the linear increase of one unit onthe x-axis corresponding to an increase of musical skill level orachievement by 0.61 units. If we expressed this in terms of anexperimental between-groups design, this rc value of 0.61 wouldtranslate to a Cohen’s d of 1.52 which implicates a very large effect(Ellis, 2010, p. 16). In our view, this is a strong argument for theeminent importance of long-term DP for skill acquisition andachievement.

FIGURE 4 | Illustration of the (linear) correlation (rc = 0.61) between

indicators of DP and musical achievement based on a simulation with

N = 788 normal distributed cases with a mean of 0. An increase of 1unit on the x-axis corresponds to an increase of 0.61 units on the y -axis.






In summary, it is incorrect to interpret our findings (rc =0.61) as evidence that DP explains 36% of the variance in attainedmusic performance. Instead, it is correct to state that the currentlytrackable correlation between an approximation of deliberatepractice with indicators such as solitary study or task-releventtraining experiences is related to measurements of music perfor-mance with rc = 0.61.

FUTURE PERSPECTIVESCurrently, there is a lack of controlled empirical studies basedon the expertise theory in the domain of music. This problemis reflected in the small number of studies (N = 13) conductedover the last 20 years which matched the rigorous selection cri-teria of our meta-analysis. One of the main challenges in thefuture will therefore be to extend the base of reliable experi-mental data. This means that studies should use state of theart measurements of relevant deliberate practice durations (e.g.,year-by-year retrospective reports, diaries etc.) and objective andreliable assessments of performance variables (e.g., preferablyhard performance measurements or consensual expert ratings ofperformance achievements). All of this was demanded many yearsago (e.g., Ericsson and Smith, 1991). The use of standardizedperformance tasks (e.g., intact performance such as sight-readingwith a pacing voice or isolated subskills such as scale playingat a given speed) with the objective measurement of perfor-mance and additional information on their reliabilities will bemandatory for investigating the “true” relationship between task-specific practice and musical achievement. This demand under-scores Ericsson’s (2014, p. 16) claim that “the expert-performanceframework restricts its research to objectively measurable perfor-mance. It rejects research based on supervisor ratings and othersocial indicators. . . .” Consequently, self-reports on abilities, therating of a musican’s skill level by an orchestra’s conductor, andreports of parents about their child’s level of achievement are notacceptable as objective indicators of performance. The questionof whether the expert performance framework generalizes to thegeneral population also awaits investigation (Ericsson, 2014). Asour findings are currently limited to music, it will be necessary tocross-validate them with meta-analytic findings in other domainsof expertise, such as sports or chess. The likelihood of their beinggeneralizable is high, though, due to the methodological rigor ofour study.

One general problem for the domain of music is that timeestimations of practice durations are only approximate indicatorsof deliberate practice, which by definition only constitutes opti-mized practice and training activities. If we were able to identifythe actual amount of deliberate practice inherent in the dura-tional estimates that currently also include suboptimal practiceactivities, especially in sub-expert populations, then the aggre-gated correlations could certainly be higher than rc = 0.61.Solitary practice might also not cover all aspects of deliberatepractice (e.g., competition experience). Thus, our figure of rc =0.61 might currently be considered as the theoretically lowerbound of the true effect of DP. The most suitable future studiesthat could untangle this empirical conundrum would includemicro-analyses of practice activities and in particular longitudi-nal studies like the one’s by McPherson et al. (2012) for music; or

Gruber et al. (1994) for chess. Such studies should be the natu-ral next step in the quest for the factors that mediate expert andexceptional performance.

AUTHOR CONTRIBUTIONSConceived and designed the meta-analysis: Andreas C. Lehmann,Reinhard Kopiez, Friedrich Platz, Anna Wolf. Conducted thesearch for references: Reinhard Kopiez, Anna Wolf, FriedrichPlatz, Andreas C. Lehmann. Analyzed the data: Friedrich Platz,Anna Wolf, Reinhard Kopiez, Andreas C. Lehmann. Wrote thepaper: Friedrich Platz, Reinhard Kopiez, Andreas C. Lehmann,Anna Wolf.

ACKNOWLEDGMENTSThis study was supported by a grant from the German ResearchFoundation (DFG Grant No. KO 1912/9-1) awarded to the secondand third author. We thank David Z. Hambrick for his very help-ful cooperation, Hans-Christian Jabusch, and Gary McPhersonfor making their data available to us.

REFERENCESBengtsson, S. L., Nagy, Z., Skare, S., Forsman, L., Forssberg, H., and Ullén, F.

(2005). Extensive piano practicing has regionally specific effects on white matterdevelopment. Nat. Neurosci. 8, 1148–1150. doi: 10.1038/nn1516

Bergee, M. J. (2003). Faculty interjudge reliability of music performance evaluation.J. Res. Music Educ. 51, 137–150. doi: 10.2307/3345847

Booth, A., Papaioannou, D., and Sutton, A. (2012). Systematic Approaches to aSuccessful Literature Review. London: Sage.

Borenstein, M. (2010). Comprehensive Meta-Analysis (2.0) [Computer software].Englewood, NJ: Biostat.

Brandler, S., and Rammsayer, T. H. (2003). Differences in mental abilitiesbetween musicians and non-musicians. Psychol. Music 31, 123–138. doi:10.1177/0305735603031002290

Campitelli, G., and Gobet, F. (2011). Deliberate practice: necessary but not suffi-cient. Curr. Dir. Psychol. Sci. 20, 280–285. doi: 10.1177/0963721411421922

Chabris, C. F. (1999). Prelude or requiem for the “Mozart effect?” Nature 400,826–827. doi: 10.1038/23608

Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences. Hillsdale, NJ:Lawrence Erlbaum.

Cohn, L. D., and Becker, B. J. (2003). How meta-analysis increases statistical power.Psychol. Methods 8, 243–253. doi: 10.1037/1082-989X.8.3.243

Cooper, H., Hedges, L. V., and Valentine, J. C. (eds.). (2009). The Handbook ofResearch Synthesis and Meta-Analysis. New York, NY: Russell Sage Foundation.

Deeks, J. J., Higgins, J. P. T., and Altman, D. G. (2008). “Analysing data andundertaking meta-analyses,” in Cochrane Handbook for Systematic Reviews ofInterventions, eds J. P. T. Higgins and S. Green (Chichester: Wiley-Blackwell),243–296. doi: 10.1002/9780470712184.ch9

Detterman, D. K. (ed.) (2014). Acquiring expertise: ability, practice, and otherinfluences [Special issue]. Intelligence 45, 1–123.

Egger, M., Smith, G. D., Schneider, M., and Minder, C. (1997). Bias in meta-analysis detected by a simple, graphical test. Br. Med. J. 315, 629–634. doi:10.1136/bmj.315.7109.629

Ellis, P. D. (2010). The Essential Guide to Effect Sizes: Statistical Power, Meta-Analysis, and the Interpretation of Research Results. Cambridge: CambridgeUniversity Press. doi: 10.1017/CBO9780511761676

Ericsson, K. A. (ed.). (1996). The Road to Excellence: the Acquisition of ExpertPerformance in the Arts and Sciences, Sports and Games. Mahwah, NJ: LawrenceErlbaum.

Ericsson, K. A. (2006). “The influence of experience and deliberate practice on thedevelopment of superior expert performance,” in The Cambridge Handbook ofExpertise and Expert Performance, eds K. A. Ericsson, N. Charness, P. J. Feltovich,and R. R. Hoffman (Cambridge: Cambridge University Press), 683–703.

Ericsson, K. A. (2014). Why expert performance is special and cannot be extrap-olated from studies of performance in the general population: A response tocriticisms. Intelligence 45, 81–103. doi: 10.1016/j.intell.2013.12.001





∗Ericsson, K. A., Krampe, R. T., and Tesch-Römer, C. (1993). The role of deliberatepractice in the acquisition of expert performance. Psychol. Rev. 100, 363–406.doi: 10.1037/0033-295X.100.3.363

Ericsson, K. A., and Lehmann, A. C. (1999). “Expertise,” in Encyclopedia ofCreativity, eds M. A. Runco and S. Pritzker (New York, NY: Academic Press),695–707.

Ericsson, K. A., and Smith, J. (1991). “Prospects and limits in the empirical studyof expertise: an introduction,” in Toward a General Theory of Expertise: Prospectsand Limits, eds K. A. Ericsson and J. Smith (Cambridge: Cambridge UniversityPress), 1–38.

Gruber, H., and Lehmann, A. C. (2008). “Entwicklung von Expertise undHochleistung in Musik und Sport [The development of expertise and supe-rior performance in music and sports],” in Angewandte Entwicklungspsychologie(Enzyklopädie der Psychologie, Vol. C/V/7), eds F. Petermann and W. Schneider(Göttingen: Hogrefe), 497–519.

Gruber, H., Renkl, A., and Schneider, W. (1994). Expertise undGedächtnisentwicklung: längsschnittliche Befunde aus der Domäne Schach[expertise and the development of memory: longitudinal results from thedomain of chess]. Zeitschrift für Entwicklungspsychologie und PädagogischePsychologie 26, 53–70.

Gulliksen, H. (1950). Theory of Mental Tests. New York, NY: Wiley and Sons. doi:10.1037/13240-000

∗Hallam, S. (1998). The predictors of achievement and dropout in instrumentaltuition. Psychol. Music 26, 116–132. doi: 10.1177/0305735698262002

Hambrick, D. Z., and Meinz, E. J. (2011). Limits on the predictive power of domain-specific experience and knowledge in skilled performance. Curr. Dir. Psychol.Sci. 20, 275–279. doi: 10.1177/0963721411422061

Hambrick, D. Z., Oswald, F. L., Altmann, E. M., Meinz, E. J., Gobet, F., andCampitelli, G. (2014). Deliberate practice: is that all it takes to become anexpert? Intelligence 45, 34–45. doi: 10.1016/j.intell.2013.04.001

Hetland, L. (2000). Listening to music enhances spatial-temporal reasoning: evi-dence for the “Mozart effect.” J. Aesthet. Educ. 34, 105–148. doi: 10.2307/3333640

Higgins, J. P. T., and Green, S. (eds.). (2008). Cochrane Handbook forSystematic Reviews of Interventions. Chichester: Wiley-Blackwell. doi:10.1002/9780470712184

Howe, M. J. A., Davidson, J. W., and Sloboda, J. A. (1998). Innate talents:reality or myths? Behav. Brain Sci. 21, 399–442. doi: 10.1017/S0140525X9800123X

Hunter, J. E., and Schmidt, F. L. (2004). Methods of Meta-Analysis: Correcting Errorand Bias in Research Findings. London: Sage.

∗Jabusch, H.-C., Alpers, H., Kopiez, R., Vauth, H., and Altenmüller, E. (2009).The influence of practice on the development of motor skills in pianists: alongitudinal study in a selected motor task. Hum. Mov. Sci. 28, 74–84. doi:10.1016/j.humov.2008.08.001

Jabusch, H.-C., Vauth, H., and Altenmüller, E. (2004). Quantification of focaldystonia in pianists using scale analysis. Mov. Disord. 19, 171–180. doi:10.1002/mds.10671

∗Jabusch, H.-C., Young, R., and Altenmüller, E. (2007). “Biographical predictorsof music-related motor skills in children pianists,” in International Symposiumon Performance Science, eds A. Williamon and D. Coimbra (Porto: AssociationEuropeìenne des Conservatoires, Acadeìmies de Musique et Musikhochschulen(AEC)), 363–368.

Kämpfe, J., Sedlmeier, P., and Renkewitz, F. (2011). The impact of backgroundmusic on adult listeners: a meta-analysis. Psychol. Music 39, 424–448. doi:10.1177/0305735610376261

Kopiez, R. (2012). “The role of replication studies and meta-analyses in thesearch of verified knowledge,” in 12th International Conference on MusicPerception and Cognition (ICMPC), 23–28 July, eds E. Cambouropoulos, C.Tsougras, P. Mavromatis, and K. Pastiadis (Thessaloniki: Aristotle University ofThessaloniki), 64–65.

Kopiez, R., Galley, N., and Lee, J. I. (2006). The advantage of beingnon-right-handed: the influence of laterality on a selected musical skill(sight reading achievement). Neuropsychologia 44, 1079–1087. doi: 10.1016/j.neuropsychologia.2005.10.023

Kopiez, R., Galley, N., and Lehmann, A. C. (2010). The relation between lateral-isation, early start of training, and amount of practice in musicians: a contri-bution to the problem of handedness classification. Laterality 15, 385–414. doi:10.1080/13576500902885975

∗Kopiez, R., Jabusch, H.-C., Galley, N., Homann, J.-C., Lehmann, A. C.,and Altenmüller, E. (2012). No disadvantage for left-handed musicians:the relationship between handedness, felt constraints and performance-related skills in pianists and string players. Psychol. Music 40, 357–384. doi:10.1177/0305735610394708

Kopiez, R., and Lee, J. I. (2006). Towards a dynamic model of skills involved in sightreading music. Music Educ. Res. 8, 97–120. doi: 10.1080/14613800600570785

∗Kopiez, R., and Lee, J. I. (2008). Towards a general model of skills involved in sightreading music. Music Educ. Res. 10, 41–62. doi: 10.1080/14613800701871363

∗Kornicke, L. E. (1992). An Exploratory Study of Individual Difference Variables inPiano Sight-Reading Achievement. PhD, Indiana University.

Krampe, R. T., and Charness, N. (2006). “Aging and expertise,” in CambridgeHandbook of Expertise and Expert Performance, eds A. K. Ericsson, N. Charness,P. J. Feltovich, and R. R. Hoffman (Cambridge: Cambridge University Press),723–742. doi: 10.1017/CBO9780511816796.040

∗Krampe, R. T., and Ericsson, K. A. (1996). Maintaining excellence: deliberate prac-tice and elite performance in young and older pianists. J. Exp. Psychol. Gen. 125,331–359. doi: 10.1037/0096-3445.125.4.331

Lehmann, A. C. (1997). “The acquisition of expertise in music: efficiency of deliber-ate practice as a moderating variable in accounting for sub-expert performance,”in Perception and Cognition of Music, eds I. Deliège and J. A. Sloboda (Hove:Psychology Press), 161–187.

Lehmann, A. C. (2005). “Musikalischer Fertigkeitserwerb (Expertisierung):Theorie und Befunde [musical skill acquisition (expertization): theories andresults],” in Musikpsychologie (Handbuch der Systematischen Musikwissenschaft,Vol. 3, eds H. De La Motte-Haber and G. Rötter (Laaber: Laaber), 568–599.

Lehmann, A. C., and Ericsson, K. A. (1993). Sight-reading ability of expert pianistsin the context of piano accompanying. Psychomusicology 12, 182–195. doi:10.1037/h0094108

∗Lehmann, A. C., and Ericsson, K. A. (1996). Performance without prepara-tion: Structure and acquisition of expert sight-reading and accompanyingperformance. Psychomusicology 15, 1–29. doi: 10.1037/h0094082

Lehmann, A. C., and Gruber, H. (2006). “Music,” in Cambridge Handbook onExpertise and Expert Performance, eds K. A. Ericsson, N. Charness, P. Feltovich,and R. R. Hoffman (Cambridge: Cambridge University Press), 457–470. doi:10.1017/CBO9780511816796.026

Masicampo, E. J., and Lalande, D. R. (2012). A peculiar prevalenceof p values just below .05. Q. J. Exp. Psychol. 65, 2271–2279. doi:10.1080/17470218.2012.711335

∗McPherson, G. E. (2005). From child to musician: skill development duringthe beginning stages of learning an instrument. Psychol. Music 33, 5–35. doi:10.1177/0305735605048012

McPherson, G. E., Davidson, J. W., and Faulkner, R. (2012). Music in Our Lives:Rethinking Musical Ability, Development and Identity. Oxford: Oxford UniversityPress. doi: 10.1093/acprof:oso/9780199579297.001.0001

∗Meinz, E. J. (2000). Experience-based attenuation of age-related differencesin music cognition tasks. Psychol. Aging 15, 297–312. doi: 10.1037/0882-77974.15.2.297

∗Meinz, E. J., and Hambrick, D. Z. (2010). Deliberate practice is necessarybut not sufficient to explain individual differences in piano sight-readingskill: the role of working memory capacity. Psychol. Sci. 21, 914–919. doi:10.1177/0956797610373933

Mishra, J. (2014). Improving sightreading accuracy: a meta-analysis. Psychol. Music42, 131–156. doi: 10.1177/0305735612463770

Nandagopal, K., and Ericsson, K. A. (2012). “Enhancing students’ performance intraditional education: Implications from the expert performance approach anddeliberate practice,” in APA Educational psychology handbook (Vol. 1, Theories,constructs, and critical issues), eds K. R. Harris, S. Graham, and U. Urdan(Washington, DC: American Psychological Association), 257–293.

Pietschnig, J., Voracek, M., and Formann, A. K. (2010). Mozart effect–Shmozarteffect: a meta-analysis. Intelligence 38, 314–323. doi: 10.1016/j.intell.2010.03.001

Platz, F., and Kopiez, R. (2012). When the eye listens: a meta-analysis of how audio-visual presentation enhances the appreciation of music performance. MusicPercept. 30, 71–83. doi: 10.1525/mp.2012.30.1.71

Rosenthal, R. (1979). The “file drawer problem” and tolerance for null results.Psychol. Bull. 86, 638–641. doi: 10.1037/0033-2909.86.3.638

Rothstein, H. R., and Hopewell, S. (2009). “Grey literature,” in The Handbook ofResearch Synthesis and Meta-Analysis, 2nd Edn., eds H. Cooper, L. V. Hedges,and J. C. Valentine (New York, NY: Russel Sage Foundation), 103–125.






Rothstein, H. R., Sutton, A. J., and Borenstein, M. (eds.). (2005). Publication Bias inMeta-Analysis: Prevention, Assessment and Adjustments. Chichester: John Wiley& Sons. doi: 10.1002/0470870168

Ruthsatz, J., Detterman, D., Griscom, W. S., and Cirullo, B. A. (2008). Becoming anexpert in the musical domain: it takes more than just practice. Intelligence 36,330–338. doi: 10.1016/j.intell.2007.08.003

Schmidt, F. L., and Le, H. A. (2005). Hunter-Schmidt Meta-Analysis Programs. V1.1. Iowa, IA.

Shavinina, L. (2009). “A unique type of representation is the essence of gifted-ness: towards a cognitive-developmental theory,” in International Handbookon Giftedness, ed L. Shavinina (Berlin: Springer Science and Business Media),231–257.

Sloboda, J. A. (2000). Individual differences in music performance. Trends Cogn.Sci. 4, 397–403. doi: 10.1016/S1364-6613(00)01531-X

∗Tuffiash, M. (2002). Predicting Individual Differences in Piano Sight-Reading Skill:Practice, Performance, and Instruction. Unpublished master’s thesis, Florida StateUniversity.

Ullén, F., Forsman, L., Blom, O., Karabanov, A., and Madison, G. (2008).Intelligence and variability in a simple timing task share neural substrates in theprefrontal white matter. J. Neurosci. 28, 4238–4243. doi: 10.1523/JNEUROSCI.0825-08.2008

Vandervert, L. R. (2009). “Working memory, the cognitive functions of the cere-bellum and the child prodigy,” in International Handbook on Giftedness, ed L.Shavinina (Berlin: Springer Science and Business Media), 295–316.

∗References marked with an asterisk indicate studies included in the meta-analysis. The in-text citations to studies selected for meta-analysis are notpreceded by asterisks.

Winner, E. (1996). “The rage to master: the decisive role of talent in the visual arts,”in The Road to Excellence: the Acquisition of Expert Performance in the Arts andSciences, Sports and Games, ed K. A. Ericsson (Mahwah, NJ: Lawrence Erlbaum),271–301.

Ziegler, A. (2009). “Research on giftedness in the 21st century,” in InternationalHandbook on Giftedness, ed L. Shavinina (Berlin: Springer Science and BusinessMedia), 1509–1524. doi: 10.1007/978-1-4020-6162-2_78


Received: 10 April 2014; accepted: 06 June 2014; published online: 26 June 2014.Citation: Platz F, Kopiez R, Lehmann AC and Wolf A (2014) The influence of delib-erate practice on musical achievement: a meta-analysis. Front. Psychol. 5:646. doi:10.3389/fpsyg.2014.00646This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Platz, Kopiez, Lehmann and Wolf. This is an open-access articledistributed under the terms of the Creative Commons Attribution License (CC BY).The use, distribution or reproduction in other forums is permitted, provided theoriginal author(s) or licensor are credited and that the original publication in thisjournal is cited, in accordance with accepted academic practice. No use, distribution orreproduction is permitted which does not comply with these terms.












GENERAL COMMENTARYpublished: 17 July 2014

doi: 10.3389/fpsyg.2014.00751

Facing facts about deliberate practiceDavid Z. Hambrick1*, Erik M. Altmann1, Frederick L. Oswald2, Elizabeth J. Meinz3 and Fernand Gobet4

1 Department of Psychology, Michigan State University, East Lansing, MI, USA2 Department of Psychology, Rice University, Houston, TX, USA3 Department of Psychology, Southern Illinois University Edwardsville, Edwardsville, IL, USA4 Institute of Psychology, Health, and Society, University of Liverpool, Liverpool, UK*Correspondence: [email protected]

Edited by:


Reviewed by:

Lena Rachel Quinto, Macquarie University, AustraliaMichael H. Connors, Macquarie University, Australia

Keywords: deliberate practice, music, expertise, expert performance, individual differences, talent

A commentary on

The influence of deliberate practice onmusical achievement: a meta-analysisby Platz, F., Kopiez, R., Lehmann, A. C., andWolf, A. (2014). Front. Psychol. 5:646. doi:10.3389/fpsyg.2014.00646

More than 20 years ago, Ericsson and col-leagues proposed that “individual differ-ences in ultimate performance can largelybe accounted for by differential amountsof past and current levels of practice”(Ericsson et al., 1993, p. 392). We empir-ically tested this claim through a meta-analysis of studies of music and chess(Hambrick et al., 2014). The claim was notsupported. Deliberate practice accountedfor about one-third of the reliable variancein performance in each domain, leavingmost of the variance explainable by otherfactors.

Focusing on music, Platz et al. (2014)identified 13 studies of the relationshipbetween deliberate practice and perfor-mance and found a correlation of 0.61after correcting for unreliability. We creditPlatz et al. for their effort and thank themfor their criticisms of our meta-analysis.However, none of these criticisms chal-lenge our conclusion that deliberate prac-tice is not as important as Ericsson andcolleagues have argued.

Platz et al.’s (2014) major criticism tar-gets our conclusion that deliberate prac-tice accounted for 30% of the variancein music performance. They write that“relationships between variables should beinterpreted in terms of linear relation-ships” (p. 10), and that “it is incorrect

to interpret our findings (rc = 0.61) asevidence that DP explains 36% of thevariance in attained music performance”(p. 11). They base this criticism on Hunterand Schmidt’s (2004) argument that effectsizes from meta-analyses (and primaryresearch) be reported as correlations ratherthan estimates of variance accounted for(i.e., rs rather than r2s).

Platz et al.’s (2014) criticism is puzzlingfor two reasons. First, other researchershave characterized the importance ofdeliberate practice in terms of vari-ance (individual differences) accountedfor—including not only Ericsson et al.(1993), but also two authors of the Platzet al. article (Reinhard Kopiez and AndreasLehmann). For example, Kopiez and col-leagues concluded that “the total life prac-tice time at the beginning of the study cor-related moderately with the baseline per-formance values and predicted only 17%of their variance” (Jabusch et al., 2009,p. 80, italics added; see also Lehmannand Ericsson, 1996; Kopiez and Lee, 2006,2008). Second, Hunter and Schmidt’s(2004) point is not that r2 is statisticallyincorrect. Indeed, r and r2 are both stan-dard indexes of effect size (Cohen, 1988),providing different ways to conceptual-ize the strength of statistical relationships.Rather, their point is that r2 can make the-oretically and practically important rela-tionships seem trivially small—as when acorrelation of, say, 0.30 between a predic-tor and an outcome is dismissed because“only” 9% of the variance is explained.For this reason, we reported both r andr2 values in our meta-analysis. Moreover,to avoid trivializing the role of deliberate

practice, we have repeatedly emphasizedits importance—the necessity of it forbecoming an expert. In no less a pub-lic forum than the opinion pages of TheNew York Times, two of us commented thatthere is no denying the “power of practice”(Hambrick and Meinz, 2011). Again, ourconclusion is not that deliberate practiceis unimportant, either statistically or the-oretically; it is that deliberate practice isnot as important as Ericsson and colleagueshave argued, in the precise sense that fac-tors other than deliberate practice accountfor most of the variance in performance.Platz et al. apparently miss this point.

Platz et al. (2014) also take aim atthe criteria we used for including a studyin our meta-analysis, calling them “intu-itive” (p. 4). In fact, our criteria were dic-tated by the theoretical claim we soughtto test and were clearly stated in ourarticle—measures of accumulated amountof deliberate practice and performancewere collected and a correlation betweenthese measures was reported. Platz et al.did find a few studies in their literaturesearch that we did not, but this does notbear on our conclusion that deliberatepractice is not as important as Ericssonand colleagues have argued. In fact, theresults of Platz et al.’s meta-analysis sup-port this conclusion: A correlation of 0.61between deliberate practice and music per-formance leaves room for two additionalorthogonal predictors of nearly the samemagnitude (rs = 0.56).

Perhaps with an inkling of this, Platzet al. (2014) argue that their correlationof 0.61 might be regarded as the “theo-retically lower bound of the true effect of

















Hambrick et al. Facing facts

DP” (p. 11) because “time estimations ofpractice durations are only approximateindicators of deliberate practice” (p. 11).But their correlation could equally wellbe regarded as an upper bound on thetrue effect of deliberate practice. For exam-ple, using retrospective questionnaires tomeasure deliberate practice could lead toinflated correlations between deliberatepractice and performance if people basepractice estimates on their skill rather thanrecollections of engaging in practice. Themore general problem with Platz et al.’sargument is that it can always be made:if the correlation between deliberate prac-tice and performance is not as high asone likes, one can always argue that this isbecause the measure of deliberate practiceis imperfect—making it impossible to fal-sify hypotheses about the predictive valueof deliberate practice.

Finally, some measures used by Platzet al. (2014) may not be estimates ofdeliberate practice. For example, for somestudies, they used the correlation betweennumber of accompanying performancesand sight-reading performance, but num-ber of accompanying performances couldbe considered a measure of what Ericssonet al. (1993) termed “work,” as distinctfrom deliberate practice. Platz et al. arealso inconsistent in what they consider theaccumulation period for deliberate prac-tice (e.g., lifetime for some studies, to age18 for others).

The bottom line is that, in all majordomains in which deliberate practice hasbeen studied, most of the variance in per-formance is explained by factors otherthan deliberate practice (Macnamara et al.,

2014). These factors may include startingage (Gobet and Campitelli, 2007), workingmemory capacity (Meinz and Hambrick,2010), and genes (Hambrick and Tucker-Drob, 2014). For scientists, the task nowis to develop and test falsifiable theoriesof expertise that include as many relevantconstructs as possible.

REFERENCESCohen, J. (1988). Statistical Power Analysis for

the Behavioral Sciences. Hillsdale, NJ: LawrenceErlbaum.

Ericsson, K. A., Krampe, R. T., and Tesch-Römer, C.(1993). The role of deliberate practice in the acqui-sition of expert performance. Psychol. Rev. 100,363–406. doi: 10.1037/0033-295X.100.3.363

Gobet, F., and Campitelli, G. (2007). The role ofdomain-specific practice, handedness, and start-ing age in chess. Dev. Psychol. 43, 159–172. doi:10.1037/0012-1649.43.1.159

Hambrick, D. Z., and Meinz, E. J. (2011). Sorry,strivers: talent matters. The New York Times 12.

Hambrick, D. Z., Oswald, F. L., Altmann, E. M.,Meinz, E. J., Gobet, F., and Campitelli, G.(2014). Deliberate practice: is that all it takesto become an expert? Intelligence 45, 34–45. doi:10.1016/j.intell.2013.04.001

Hambrick, D. Z., and Tucker-Drob, E. (2014). Thegenetics of music accomplishment: evidence forgene-environment correlation and interaction.Psychon. Bull. Rev. doi: 10.3758/s13423-014-0671-9. [Epub ahead of print].

Hunter, J. E., and Schmidt, F. L. (2004). Methodsof Meta-Analysis: Correcting Error and Bias inResearch Findings. Thousand Oaks, CA: Sage.

Jabusch, H. C., Alpers, H., Kopiez, R., Vauth,H., and Altenmuller, E. (2009). The influenceof practice on the development of motorskillsin pianists: a longitudinal study in a selectedmotor task. Hum. Mov. Sci. 28, 74–84. doi:10.1016/j.humov.2008.08.001

Kopiez, R., and Lee, J. I. (2006). Towards adynamic model of skills involved in sightreading music. Music Educ. Res. 8, 97–120. doi:10.1080/14613800600570785

Kopiez, R., and Lee, J. I. (2008). Towards a gen-eral model of skills involved in sight read-ing music. Music Educ. Res. 10, 41–62. doi:10.1080/14613800701871363

Lehmann, A. C., and Ericsson, K. A. (1996).Performance without preparation: structure andacquisition of expert sight-reading and accom-panying performance. Psychomusicology 15, 1–29.doi: 10.1037/h0094082

Macnamara, B. N., Hambrick, D. Z., and Oswald, F.L. (2014). Deliberate practice and performance inmusic, games, sports, education, and professions:a meta-analysis. Psychol. Sci. doi: 10.1177/0956797614535810. [Epub ahead of print].

Meinz, E. J., and Hambrick, D. Z. (2010). Deliberatepractice is necessary but not sufficient to explainindividual differences in piano sight-readingskill: the role of working memory capacity.Psychol. Sci. 21, 914–919. doi: 10.1177/0956797610373933

Platz, F., Kopiez, R., Lehmann, A. C., and Wolf,A. (2014). The influence of deliberate prac-tice on musical achievement: a meta-analysis.Front. Psychol. 5:646. doi: 10.3389/fpsyg.2014.00646


Received: 20 June 2014; accepted: 27 June 2014;published online: 17 July 2014.Citation: Hambrick DZ, Altmann EM, Oswald FL,Meinz EJ and Gobet F (2014) Facing facts about delib-erate practice. Front. Psychol. 5:751. doi: 10.3389/fpsyg.2014.00751This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Hambrick, Altmann, Oswald,Meinz and Gobet. This is an open-access article dis-tributed under the terms of the Creative CommonsAttribution License (CC BY). The use, distribution orreproduction in other forums is permitted, provided theoriginal author(s) or licensor are credited and that theoriginal publication in this journal is cited, in accor-dance with accepted academic practice. No use, distribu-tion or reproduction is permitted which does not complywith these terms.













OPINION ARTICLEpublished: 19 February 2014

doi: 10.3389/fpsyg.2014.00131

Training principles to advance expertiseAlice F. Healy1*, James A. Kole2 and Lyle E. Bourne Jr.1

1 Department of Psychology and Neuroscience, University of Colorado, Boulder, CO, USA2 School of Psychological Sciences, University of Northern Colorado, Greeley, CO, USA*Correspondence: [email protected]

Edited by:


Reviewed by:

Craig Speelman, Edith Cowan University, Australia

Keywords: training, acquisition, retention, transfer, efficiency, durability, generalizability, expertise

INTRODUCTIONThere are three forms of task engage-ment that are the basis of successful train-ing for expertise—acquisition, retention,and transfer—and three correspondinggoals of training—efficiency, durability,and generalizability. This paper reviewstraining conditions that facilitate acqui-sition, enhance retention, and promotetransfer to contexts not encountered dur-ing training. Diligent practice under thesetraining conditions can lead eventuallyto an elite level of performance or toexpertise.

In this review of training principles,we emphasize those that are derived fromwork in our laboratory. We have foundthat developing training that optimizesefficiency, durability, and generalizability,however, is something of a balancing actbecause what promotes efficient train-ing often comes at a price in durability,and durable training is not always gen-eralizable (see Healy et al., 2012; Bourneand Healy, 2014). These tradeoffs are duein part to the fact that training mightinvolve two different types of knowledge—declarative and procedural. Declarativeknowledge consists of facts, whereas pro-cedural knowledge consists of skills, orways to use the facts, and these two typesof knowledge differ in terms of their dura-bility and generalizability.

ACQUISITIONThere are training principles that can beused to improve the efficiency of acquiringknowledge and skills. One way to increasetraining speed involves the scheduling offeedback given to the trainee. Trial-by-trial feedback can facilitate learning inmany situations, especially by improving

motivation and by identifying and cor-recting errors. Frequent feedback can bedistracting, however, when the traineesalready have a good sense of how wellthey are responding. In those cases peri-odic summary feedback, which is providedon only some percentage of training tri-als, can be a more effective procedure forpromoting skill acquisition (Schmidt et al.,1989, demonstrated this principle with aballistic timing task).

In terms of the difficulty of new mate-rial that is presented during training, thereis an optimal zone of learnability, implyingthat training changes should be neither toodifficult nor too easy. Specifically, train-ing trials should contain information thatis a little beyond what the trainee alreadyknows or should require a bit faster ormore accurate performance (Wolfe et al.,1998).

Training can be based on mental, asopposed to physical, practice. Althoughphysical practice is better than mentalpractice in many circumstances, mentalpractice can be superior to physical prac-tice when the practice involves differ-ent effectors from those used at testing(Wohldmann et al., 2008a). For example,training with the right hand and testingwith the left hand can lead to interferenceduring physical practice, but not duringmental practice, which involves a moreabstract representation of the skill thatappears to be effector independent. Mentalpractice is especially useful, of course,when circumstances do not allow physicalpractice.

The focus of attention can be variedin training, and training usually benefitsfrom an external focus on the results ofbody movements as opposed to an internal

focus on the body movements themselves(Wulf, 2007). For example, to perfect theskill of dart throwing, attention should befocused on the trajectory of the dart ratherthan on the motion of the arm (Lohseet al., 2010).

Research has concluded that for learn-ing new skills (but not necessarily forlearning new facts) rest intervals should beinterpolated between practice trials (dis-tributed practice) rather then practicingwithout rest intervals (massed practice),especially if testing occurs after a delay fol-lowing practice (see Cepeda et al., 2006,for a review of effects involving spacing ofpractice).

When engaged in prolonged work ona routine task, it is often advisable tointroduce into the training regimen a cog-nitive antidote, which is a simple cogni-tive requirement that can be added tominimize task disengagement and bore-dom and to mitigate a decline in accu-racy across trials (Kole et al., 2008). Forexample, when entering into a computera long sequence of numbers, alternat-ing between using the + and − keysas the concluding keystroke, rather thanusing only a single key to conclude eachnumber, increases the accuracy of digitentry and eliminates the usual increasein errors that occurs with increasingpractice.

Task requirements can sometimes bebroken up into parts, with practice atthe beginning of training involving onlyone or more parts of a task rather thanthe whole task (Wickens et al., 2012).The recommendation to use part-tasktraining applies to a segmented task,when the parts are performed sequen-tially (Wightman and Lintern, 1985), but













Healy et al. Training principles to advance expertise

not to a fractionated task, when the partsare performed simultaneously (Adams andHufford, 1962).

RETENTIONThe durability of training can be enhancedby selected training principles. Some ofthese principles benefit training retentionwith a cost in training efficiency, but oth-ers benefit retention with little or no costin training efficiency.

RETENTION VARIABLES WITH A COST INEFFICIENCYComplications can sometimes be added totraining to promote retention (Schneideret al., 2002). However, not all complica-tions that increase task difficulty are desir-able (Bjork, 1994); difficulties at trainingfacilitate retention only when they forcelearners to apply task-relevant cognitiveprocesses (McDaniel and Butler, 2011).For example, in training RADAR detec-tion, adding to training a secondary taskthat was irrelevant lowered RADAR detec-tion performance at test, but adding arelevant secondary task at training aidedRADAR detection accuracy at test (Younget al., 2011).

As research on experts has shown(Ericsson et al., 1980), training shouldmake strategic use of knowledge thattrainees already have (Kole and Healy,2007). For example, when learning factsabout unknown people, associating eachof those people with someone famil-iar (e.g., a friend or relative) requiresadditional learning but can considerablyenhance acquisition and retention of thosefacts.

Also following the methods of experts,training can be improved by using mem-ory strategies, such as the keyword method,by which two items can be associatedin memory through a keyword servingas a mediating link (Kole and Healy,2013). When learning the associationbetween the French word jambe and theEnglish word leg, for example, learn-ers could use the keyword jam andform an interactive image of some stickyjam on someone’s leg. This method willallow the learners to derive the transla-tion of jambe, but it does require form-ing extra links from jambe to jam andfrom jam to leg, which slows initialacquisition.

RETENTION VARIABLES WITH NO COST INEFFICIENCYItems can be grouped together in chunksduring training to promote retention. Itemchunking yields no loss, and probably again, in training efficiency (Miller, 1956).For example, when required to type afour-digit number (e.g., 9632), individu-als often find it helpful to represent thenumber as two two-digit chunks (96 and32) (Fendrich et al., 1991). Also, mem-ory experts can learn a long list of digitsby dividing them into chunks represent-ing familiar sequences like running timesor dates (Ericsson et al., 1980).

Similar to chunking, two separate taskscan often be combined into a single func-tional task to promote durability with nocost to efficiency. If a secondary task iscombined with a primary task at test,the two tasks should be combined duringtraining as well for best test performance(Wohldmann et al., 2012). In fact, in somecases removing at test a difficult secondarytask that was available during trainingcan retard test performance; for example,removing an alphabet-counting task usedduring the training of time estimation canincrease errors in time estimation duringtest (Healy et al., 2005).

For optimal retention, the proceduresused in training should be duplicated attest. Procedural reinstatement is effectivebecause declarative information (facts)shows strong generalizability but weakdurability, whereas procedural informa-tion (skills) shows strong durability butweak generalizability (Lohse and Healy,2012). Despite the high degree of transferfor declarative information, when learn-ing facts it is best to make sure that thereis an overlap in the processing requiredat training and test (i.e., ensure that thereis transfer appropriate processing; Morriset al., 1977; Roediger et al., 1989). Anotherway to improve the retention of factualmaterial is by studying it at its most mean-ingful level, as opposed to studying it at asuperficial level (e.g., paying attention tothe sound or spelling of the material), aneffect called depth of processing (Craik andLockhart, 1972).

Practice retrieving factual information,instead of simply restudying the mate-rial, can improve retention, demonstrat-ing a testing effect (Roediger and Karpicke,2006). A related method to enhance fact

retention is to generate the informationrather than just to read it, demonstrat-ing a generation effect (Slamecka and Graf,1978). For example, if the task is toremember the answers to a set of multi-plication problems, it is better to generateor verify the answers to the problems thanto simply read them or perform the multi-plication using a calculator (Crutcher andHealy, 1989). See Karpicke and Zaromb(2010) for a comparison of the similarbut not identical benefits of testing andgeneration.

For skills, periodic restudy of mate-rial during periods of disuse, calledrefresher training, enhances skill reten-tion (McDaniel, 2012). On the otherhand, training should not be done morethan necessary, resulting in overlearning orovertraining, because extra practice pro-duces diminishing performance enhance-ment returns (Driskell et al., 1992).

TRANSFERImproving the generalizabilty of knowl-edge and skills is considerably more dif-ficult than improving their efficiency anddurability. There are only a few train-ing principles shown to be effective forenhancing task transfer.

One way to enhance transfer is tochange the conditions of practice period-ically, thereby increasing the variability ofpractice (Schmidt and Bjork, 1992). Notall forms of variable practice are effec-tive, however. Varying parameters withina single generalized motor program canenhance transfer, but varying the gener-alized motor programs themselves mightnot benefit transfer. For example, learn-ing how to cope with a variety of defectivecomputer mice that vary in their prop-erties (e.g., a mouse that reverses onlyvertical movements and another mousethat reverses both horizontal and verti-cal movements) does not enhance transferto a new type of defective mouse (e.g., amouse that reverses only horizontal move-ments) (Healy et al., 2006). Practicing tomove a single defective mouse to a varietyof targets, however, does enhance transferto moving that same mouse to new targets(Wohldmann et al., 2008b).

For tasks involving quantitative estima-tion (e.g., of country populations or inter-city distances), a technique of seeding theknowledge base can be used, in which a few






selected examples are given that define adomain and thereby enhance transfer tounpracticed examples (Brown and Siegler,1996).

When possible, the discovery of rulescan lead to generalizable knowledge, sothat searching for systematic relationshipsamong examples can bolster transfer oftraining, showing the advantage for rulesvs. instance memory (Bourne et al., 2010;McDaniel et al., in press). For example, inlearning how to choose between the alter-nate pronunciations for the word the (asthuh or thee) individuals can either learnthe pronunciation preceding every singleword or can learn the rule that thuh is usedpreceding words starting with a consonantsound and thee is used preceding wordsstarting with a vowel sound.

CONCLUDING REMARKSThis summary, which is outlined inTable 1, does not include all trainingprinciples that have been shown to beeffective in promoting training efficiency,durability, and generalizability. For addi-tional principles, readers are referred to

Table 1 | Training principles as a function of

form of task engagement and training goal.

ACQUISITION, EFFICIENCY

Scheduling of feedbackZone of learnabilityMental practiceFocus of attentionSpacing of practiceCognitive antidotePart-task trainingRETENTION, DURABILITY

Retention variables with a cost in efficiencyTask difficultyStrategic use of knowledgeMemory strategiesRetention variables with no cost in efficiencyItem chunkingFunctional taskProcedural reinstatementDepth of processingTesting effectGeneration effectRefresher trainingOverlearning or overtrainingTRANSFER, GENERALIZABILITY

Variability of practiceSeeding the knowledge base

Rules vs. instance memory

Bourne and Healy (2014) and Healy et al.(2012), and for an initial attempt toaccount for the principles in a single the-oretical framework, readers are referred toJones et al. (2012), which is available byrequest from the corresponding author.

Almost all of the studies reviewedhere have involved novice learners.Nevertheless, these same principles shouldapply for training at any level, so theyshould be informative with respect to theissue of training to an elite level of perfor-mance. In fact, some of these principles(e.g., focus of attention) seem to applymore strongly for experts than for novices(e.g., Perkins-Ceccato et al., 2003). Toreach elite levels of performance, how-ever, the learners need to couple thesetraining principles with the use of deliber-ate, highly focused, and effortful practice(Ericsson et al., 1993).

AUTHOR CONTRIBUTIONSAlice F. Healy wrote the initial and finaldrafts of the manuscript. James A. Koleand Lyle E. Bourne Jr. edited the initialdraft of the manuscript and suggested revi-sions of it.

ACKNOWLEDGMENTSWe are indebted to the members ofthe Center for Research on Trainingat the University of Colorado fortheir helpful suggestions about thisresearch. The research reported herewas supported by Army Research OfficeGrant W911NF-05-1-0153 and NationalAeronautics and Space AdministrationGrant NNX10AC87A.

REFERENCESAdams, J. A., and Hufford, L. E. (1962). Contributions

of a part-task trainer to the learning and relearningof a time-shared flight maneuver. Hum. Factors 4,159–170.

Bjork, R. A. (1994). “Memory and metamemoryconsiderations in the training of human beings,”

in Metacognition: Knowing About Knowing, edsJ. Metcalfe and A. Shimamura (Cambridge, MA:MIT Press), 185–205.

Bourne, L. E. Jr., and Healy, A. F. (2014). TrainYour Mind for Peak Performance: A Science-BasedApproach for Achieving Your Goals. Washington,DC: American Psychological Association.

Bourne, L. E. Jr., Raymond, W. D., and Healy, A. F.(2010). Strategy selection and use during classifica-

tion skill acquisition. J. Exp. Psychol. Learn. Mem.

Cogn. 36, 500–514. doi: 10.1037/a0018599Brown, N. R., and Siegler, R. S. (1996). Long-

term benefits of seeding the knowledge base.

Psychol. Bull. Rev. 3, 385–388. doi: 10.3758/BF03210766

Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., andRohrer, D. (2006). Distributed practice in verbalrecall tasks: a review and quantitative synthesis.Psychol. Bull. 132, 354–380. doi: 10.1037/0033-2909.132.3.354

Craik, F. I. M., and Lockhart, R. S. (1972). Levelsof processing: a framework for memory research.

J. Verbal Learn. Verbal Behav. 11, 671–684. doi:10.1016/S0022-5371(72)80001-X

Crutcher, R. J., and Healy, A. F. (1989). Cognitiveoperations and the generation effect. J. Exp.Psychol. Learn. Mem. Cogn. 15, 669–675. doi:10.1037/0278-7393.15.4.669

Driskell, J. E., Willis, R. P., and Copper, C. (1992).Effect of overlearning on retention. J. Appl. Psychol.77, 615–622. doi: 10.1037/0021-9010.77.5.615

Ericsson, K. A., Chase, W. G., and Faloon, S.(1980). Acquisition of a memory skill. Science 208,1181–1182. doi: 10.1126/science.7375930


Fendrich, D. W., Healy, A. F., and Bourne, L. E. Jr.(1991). Long-term repetition effects for motoricand perceptual procedures. J. Exp. Psychol. Learn.Mem. Cogn. 17, 137–151. doi: 10.1037/0278-7393.17.1.137

Healy, A. F., Schneider, V. I., and Bourne, L. E.Jr. (2012). “Empirically valid principles of train-ing,” in Training Cognition: Optimizing Efficiency,Durability, and Generalizability, eds A. F. Healy andL. E. Jr. Bourne (New York, NY: Psychology Press),13–39.

Healy, A. F., Wohldmann, E. L., Parker, J. T., andBourne, L. E. Jr. (2005). Skill training, reten-tion, and transfer: the effects of a concurrentsecondary task. Mem. Cogn. 33, 1457–1471. doi:10.3758/BF03193378

Healy, A. F., Wohldmann, E. L., Sutton, E. M., andBourne, L. E. Jr. (2006). Specificity effects in train-ing and transfer of speeded responses. J. Exp.Psychol. Learn. Mem. Cogn. 32, 534–546. doi:10.1037/0278-7393.32.3.534

Jones, M., Bourne, L. E. Jr., and Healy, A. F.(2012). “A compact mathematical model for pre-dicting the effectiveness of training,” in TrainingCognition: Optimizing Efficiency, Durability, andGeneralizability, eds A. F. Healy and L. E. Jr. Bourne(New York, NY: Psychology Press), 247–266.

Karpicke, J. D., and Zaromb, F. M. (2010). Retrievalmode distinguishes the testing effect from the gen-eration effect. J. Mem. Lang. 62, 227–239. doi:10.1016/j.jml.2009.11.010

Kole, J. A., and Healy, A. F. (2007). Using priorknowledge to minimize interference when learn-ing large amounts of information. Mem. Cognit.35, 124–137. doi: 10.3758/BF03195949

Kole, J. A., and Healy, A. F. (2013). Is retrievalmediated after repeated testing? J. Exp. Psychol.Learn. Mem. Cogn. 39, 462–472. doi: 10.1037/a0028880

Kole, J. A., Healy, A. F., and Bourne, L. E. Jr.(2008). Cognitive complications moderate thespeed-accuracy tradeoff in data entry: a cogni-tive antidote to inhibition. Appl. Cogn. Psychol. 22,917–937. doi: 10.1002/acp.1401





Lohse, K. R., and Healy, A. F. (2012). Exploring thecontributions of declarative and procedural infor-mation to training: a test of the procedural rein-statement principle. J. Appl. Res. Mem. Cogn. 1,65–72. doi: 10.1016/j.jarmac.2012.02.002

Lohse, K. R., Sherwood, D. E., and Healy, A. F. (2010).How changing the focus of attention affects per-formance, kinematics, and electromyography indart throwing. Hum. Mov. Sci. 29, 542–555. doi:10.1016/j.humov.2010.05.001

McDaniel, M. A. (2012). “Put the SPRINT in knowl-edge training: training with SPacing, Retrieval, andINTerleaving,” in Training Cognition: OptimizingEfficiency, Durability, and Generalizability, eds A.F. Healy and L. E. Jr. Bourne (New York, NY:Psychology Press), 267–286.

McDaniel, M. A., and Butler, A. C. (2011). “A con-textual framework for understanding when diffi-culties are desirable,” in Successful Rememberingand Successful Forgetting: A Festschrift in Honor ofRobert A. Bjork, ed A. S. Benjamin (New York, NY:Psychology Press), 175–198.

McDaniel, M. A., Cahill, M. J., Robbins, M., andWiener, C. (in press). Individual differences inlearning and transfer: stable tendencies for learn-ing exemplars versus abstracting rules. J. Exp.Psychol. Gen. doi: 10.1037/a0032963

Miller, G. A. (1956). The magical number seven, plusor minus two: some limits on our capacity for pro-cessing information. Psychol. Rev. 63, 81–97. doi:10.1037/h0043158

Morris, C. D., Bransford, J. D., and Franks, J. J.(1977). Levels of processing versus transfer appro-priate processing. J. Verbal Learn. Verbal Behav. 16,519–533. doi: 10.1016/S0022-5371(77)80016-9

Perkins-Ceccato, N., Passmore, S. R., and Lee, T.D. (2003). Effects of focus of attention dependon golfers’ skill. J. Sport Sci. 21, 593–600. doi:10.1080/0264041031000101980

Roediger, H. L. III., and Karpicke, J. D. (2006).The power of testing memory: basic research

and implications for educational practice. Perspect.Psychol. Sci. 1, 181–210. doi: 10.1111/j.1745-6916.2006.00012.x

Roediger, H. L. III., Weldon, M. S., and Challis, B. H.(1989). “Explaining dissociations between implicitand explicit measures of retention: a processingaccount,” in Varieties of Memory and Consciousness:Essays in Honour of Endel Tulving, eds H. L.III. Roediger and F. I. M. Craik (Hillsdale, NJ:Erlbaum), 3–41.

Schmidt, R. A., and Bjork, R. A. (1992). New con-ceptualizations of practice: common principles inthree paradigms suggest new concepts for train-ing. Psychol. Sci. 3, 207–217. doi: 10.1111/j.1467-9280.1992.tb00029.x

Schmidt, R. A., Young, D. E., Swinnen, S., andShapiro, D. C. (1989). Summary knowledge ofresults for skill acquisition: support for the guid-ance hypothesis. J. Exp. Psychol. Learn. Mem. Cogn.15, 352–359. doi: 10.1037/0278-7393.15.2.352

Schneider, V. I., Healy, A. F., and Bourne, L. E.Jr. (2002). What is learned under difficult con-ditions is hard to forget: contextual interferenceeffects in foreign vocabulary acquisition, reten-tion, and transfer. J. Mem. Lang. 46, 419–440. doi:10.1006/jmla.2001.2813

Slamecka, N. J., and Graf, P. (1978). The genera-tion effect: delineation of a phenomenon. J. Exp.Psychol. Hum. Learn. Mem. 4, 592–604. doi:10.1037/0278-7393.4.6.592

Wickens, C., Hutchins, S., Carolan, T., andCumming, J. (2012). “Attention and cognitiveresource load in training strategies,” in TrainingCognition: Optimizing Efficiency, Durability, andGeneralizability, eds A. F. Healy and L. E. Jr.Bourne (New York, NY: Psychology Press), 67–88.

Wightman, D., and Lintern, G. (1985). Part-task train-ing for tracking and manual control. Hum. Factors27, 267–283.

Wohldmann, E. L., Healy, A. F., and Bourne, L. E.Jr. (2008a). A mental practice superiority effect:

less retroactive interference and more transferthan physical practice. J. Exp. Psychol. Learn.Mem. Cogn. 34, 823–833. doi: 10.1037/0278-7393.34.4.823

Wohldmann, E. L., Healy, A. F., and Bourne, L.E. Jr. (2008b). Global inhibition and midcoursecorrections in speeded aiming. Mem. Cognit. 36,1228–1235. doi: 10.3758/MC.36.7.1228

Wohldmann, E. L., Healy, A. F., and Bourne, L.E. Jr. (2012). Specificity and transfer effects intime production skill: examining the role of atten-tion. Atten. Percept. Psychophys. 74, 766–778. doi:10.3758/s13414-012-0272-5

Wolfe, M. B. W., Schreiner, M. E., Rehder, B., Laham,D., Foltz, P. W., Kintsch, W., et al. (1998). Learningfrom text: matching readers and text by latentsemantic analysis. Discourse Process. 25, 309–336.doi: 10.1080/01638539809545030

Wulf, G. (2007). Attention and Motor Skill Learning.Champaign, IL: Human Kinetics.

Young, M. D., Healy, A. F., Gonzalez, C., Dutt, V.,and Bourne, L. E. Jr. (2011). Effects of trainingwith added difficulties on RADAR detection. Appl.Cogn. Psychol. 25, 395–407. doi: 10.1002/acp.1706

Received: 09 January 2014; accepted: 30 January 2014;published online: 19 February 2014.Citation: Healy AF, Kole JA and Bourne LE Jr. (2014)Training principles to advance expertise. Front. Psychol.5:131. doi: 10.3389/fpsyg.2014.00131This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Healy, Kole and Bourne. This is anopen-access article distributed under the terms of theCreative Commons Attribution License (CC BY). Theuse, distribution or reproduction in other forums is per-mitted, provided the original author(s) or licensor arecredited and that the original publication in this journalis cited, in accordance with accepted academic practice.No use, distribution or reproduction is permitted whichdoes not comply with these terms.













OPINION ARTICLEpublished: 12 June 2014

doi: 10.3389/fpsyg.2014.00580

The acquisition of expertise in the classroom: are currentmodels of education appropriate?Craig P. Speelman*

School of Psychology and Social Science, Cognition Research Group, Edith Cowan University, Joondalup, Australia*Correspondence: [email protected]



Keywords: expertise, skill, numeracy, mathematics, computer games

INTRODUCTIONThe study of expertise tends to focuson humans who can perform extraordi-nary feats. Although the way in whichexpertise is acquired is often character-ized as similar to everyday skill acquisi-tion, the attainment of basic numeracyskills is rarely considered in the same con-text as the attainment of expertise. It isclear, though, that average numeracy skillspossess all the hallmarks of expert per-formance. In this paper I argue that thetraditional classroom of Western educa-tion systems pays insufficient attention tothe idea that effective numeracy skills rep-resent a level of expertise that requiresa particular form of training. Using thefive principles of skill acquisition identifiedby Speelman and Kirsner (2005), I arguethat the modern classroom is not the mostappropriate environment for acquiringimportant cognitive skills, and that com-puter programs, such as games and tai-lored training tasks, should be considereda valuable adjunct to traditional didacticinstruction.

THE NATURE OF EXPERTISEExperts in most fields are characterizedas people who have more knowledge andabilities than non-experts. In cognitivescience, this superiority in knowledgeand abilities has been reported as bet-ter memory, more extensive knowledge,and highly developed procedural andperceptual skills. More often than not,experts are described as people who canperform extraordinary feats of mem-ory and cognition, whilst possessingthe same cognitive apparatus as non-experts. Their superior performance istypically explained as the end prod-uct of skill acquisition through many

years of dedicated practice in theirfield.

ACQUISITION OF EXPERTISEBased on over 100 years of research on skillacquisition and expertise (e.g., Anderson,1982; Logan, 1988; Ericsson et al., 1993),Speelman and Kirsner (2005) identifiedfive principles of skill acquisition thatexplain how expertise is associated withsuperior feats of cognition. In brief, thefive principles of skill acquisition statethat (1) practice leads to faster and (2)more efficient uses of knowledge, whichenables faster performance and (3) resultsin less demand on mental resources. As aresult, the performance of low level tasksbecomes second nature, and (4) this freesup mental resources that can be utilized toattempt higher level behaviors. Ultimately(5) skilled performance reflects the devel-opment of many component processes.

ACQUISITION OF NUMERACY SKILLSThe five principles of skill acquisitionwere derived from research in a broadrange of fields in which people have beenreported to have acquired expertise. As aresult, Speelman and Kirsner (2005) claimthat these principles apply to any areaof cognition in which practice can leadto improved performance. Importantly,these principles can also explain perfor-mance improvements along the trajec-tory of expertise attainment, even at levelsbelow which someone would be consid-ered an expert. So, whereas in a discus-sion of mathematics experts Butterworth(2006) considered only those 1 in sev-eral million people with extreme abilities,the principles suggest we all show ele-ments of expertise as we learn to countand add. I have been surprised to learn,

however, how little of this view of skillacquisition is reflected in current educa-tion practice. In particular, the attainmentof basic numeracy skills is rarely consid-ered as a form of expertise acquisition,and nor are difficulties with the learningof numeracy skills seen as problems inthe acquisition of expertise (e.g., a lack ofpractice). Instead, learning difficulties areoften seen as resulting from some neu-rological or developmental disorder thatadversely affects a child’s ability to learnmathematics (Clark et al., 2014; Haaseet al., 2014), or some systemic issue relatedto the school system (Ramsden, 1984;Biggs, 1999; van Kraayenoord and Elkins,2004) or the child’s culture (Whitburn,1996) or SES (OECD, 2013). And yet,numeracy skills, particularly those relatedto early learning of arithmetic and num-ber facts, share features with many exam-ples of expertise. For example, they requiremany years of practice to attain, perfor-mance relies on pattern recognition of alarge number of items (i.e., numbers andsymbols) and retrieval from memory ofa large number of facts, and expertiseattainment is characterized as a transitionthrough a hierarchy of skills (e.g., count-ing to addition) that can only occur whenperformance at the lower level has attaineda particular level of skill (Pellegrino andGoldman, 1987; Neumann et al., 2013).These are all features that have been identi-fied in experts in a range of other domains(e.g., Ericsson, 1996). It is rare to seethe development of basic numeracy skillscharacterized in this manner in educationresearch (e.g., Griffin, 2009; Lei et al., 2009;Yelland and Kildery, 2010; Neumann et al.,2013), and certainly my conversations withprimary school teachers reveals they areunaware of such concepts.












Speelman Expertise acquisition in the classroom

THE TRADITIONAL WESTERNCLASSROOMThe traditional model of the westernclassroom, with a teacher at the frontdelivering instruction to a class of 20–30children, does not sit easily with the viewof skill acquisition reflected in the fiveprinciples. Although a teacher may be ableto effectively convey declarative and pro-cedural information relevant to an arith-metic task (e.g., demonstrate how 2 × 4 isequivalent to 4 + 4), for a child to con-vert this sort of knowledge into a formof expertise requires a great deal of prac-tice. Traditionally a child would get thispractice by attempting to solve many prob-lems like 2 × 4 = ? until the teacher issatisfied that most children have graspedthe concept. But “grasping” the conceptmay not be sufficient. According to thefive principles, what counts as “sufficient”practice is determined by what comes nextin the trajectory of skill acquisition. Forinstance, if the “2×” problems are intro-duced as an extension of addition facts, itis important that the child is sufficientlyskilled at retrieving relevant addition facts.If they stumble over the idea of 4 + 4 (i.e.,they do not retrieve the answer quickly)while the teacher is explaining the “2×”concept, they may be forced to generatethe answer by a counting method (e.g.,start from 4 and count a further 4 places)(Pellegrino and Goldman, 1987). Such acounting strategy will tax working mem-ory and as a result the child may not havesufficient working memory capacity to fol-low the explanation of the “2×” concept.This strategy will also take extra time andso the student may fall behind the teacher’sexplanation. It is difficult for a teacher toensure that the pace of a lesson matches thelearning rates of all children in the class,and also to monitor that all children havemastered a concept prior to introducingthe next concept. Children who are fastlearners may become bored and disruptiveif the lesson is paced too slow. Childrenwho are slower learners may be left behindby a lesson that is paced to match the learn-ing rate of the faster learners. It would nottake too many experiences of being leftbehind for someone to believe they justdo not have a “maths brain” (Swan, 2004),or even develop a “maths phobia” (Furnerand Berman, 2003). Even if the attainmentof proficiency is assumed to occur through

homework drills, it is likely that only accu-racy is checked by a teacher, when speed ofaccess to number facts is also necessary toavoid the problem just described.

THE DECLINE IN NUMERACY SKILLSThe suggestion that the traditionalclassroom may lead to some childrenstruggling with the acquisition of basicnumeracy skills goes some way to explainthe slide in numeracy skills in many west-ern countries like Australia [as indicated inthe PISA results of 2009 (OECD, 2013)].The problem in Australia appears to belongstanding as 53% of the Australianadult population is functionally innu-merate (ABS, 2006), which indicates thatmany would not comprehend a bank state-ment. Failures in elementary mathematicscourses are likely to be compounded by alack of confidence regarding higher levelmathematics, and so many people neglectmaths at the higher levels, leading to uni-versities having to provide remedial classesfor their commencing students (Healyet al., 2010; Slattery, 2010; Arlington, 2012;Maslen, 2012).

AN ALTERNATIVE EDUCATION MODELAccording to the five principles of skillacquisition, overcoming the problemsidentified here with the traditionalWestern classroom would require childrento be presented with a structured learningprogram that involved instruction regard-ing each level of a hierarchy of concepts,interspersed with practice opportunities.Further, each child would only be intro-duced to the next level of a concept whenthey have reached some degree of fluencywith the previous level. Until that pointthey would continue practicing with prob-lems at the previous level, possibly withsome form of intervention by a tutor toensure their understanding of the conceptis appropriate. This is probably the aim ofmost teachers, however the level of mon-itoring required to ensure each child hasreached the requisite level of fluency ispossibly beyond the capacity of a teacherresponsible for 20–30 children in the oneclassroom. An alternative model would beto develop computer software in the formof games and tailored training tasks. Suchsoftware can be developed to not only pro-vide hours of practice opportunities, butit can do so in an exciting and enjoyable

manner that will hold the attention ofchildren and provide them with the moti-vation to spend many hours mastering aconcept (Rosas et al., 2003). Further, thesoftware can be designed to deliver feed-back on every response, and monitor thelevel of performance (i.e., both accuracyand response time) such that a child will beallowed to move to the next higher level ofthe concept when they have mastered theprevious level, as is the case with computergames designed purely for entertainment(Towne et al., 2014). A recent study (Mainand O’Rourke, 2011) demonstrated thebenefits of such software, where the speedand accuracy of performance on a stan-dard arithmetic test was improved forchildren who had played a maths game(Dr Kawashima’s Brain Training) on ahand held games console compared tochildren who received standard classroomlessons.

Another computer game, Numbeat, hasbeen designed explicitly according to thefive principles. This game was developedto facilitate the acquisition of basic arith-metic skills in primary school children.Numbeat requires a player to destroy some“bad” characters on the screen, beforethey convert “good” characters into “bad”characters, by filling up some destructiondevice (e.g., a cannon) with an amountof ammunition that matches the num-ber of “bad” characters. To achieve thisaim requires the player to perform severaltypes of mental arithmetic operations. Thegame is structured so that performancespeed is important (e.g., levels have timelimits). If a player beats the time limitfor a particular level, they are consideredto be sufficiently fluent with the arith-metic operations tested in that level, andso are allowed to progress to the next levelof the game, which typically represents aslightly more advanced level of arithmetic.A level is repeated if the player does notmeet the required performance standard.As such, the game approximates the delib-erate practice of challenging tasks that isrequired to acquire expertise in a domain(Towne et al., 2014). In preliminary tri-als involving 248 children, playing thisgame for 10–15 min per day for 8–10 daysresulted in an average 5% reduction in thetime to solve simple arithmetic problemspresented in a traditional manner (e.g.,3 × 4 = ?) (Speelman, 2013).





Speelman Expertise acquisition in the classroom

Although I do not claim that gamessuch as Dr Kawashima’s Brain Training andNumbeat will be the solution to all of themathematical ills facing many countries,evidence that playing such games can facil-itate the acquisition of arithmetic skills inprimary school children is a positive steptoward addressing learning difficulties inmathematics. This type of research maypave the way toward a situation whereteachers can provide the necessary con-tent lessons, and computers can facilitatethe necessary practice that will enable stu-dents to master each level of a skill beforetackling the next step in the skill hierar-chy. Ultimately such evidence supports theargument that education in basic numer-acy skills should reflect principles of exper-tise attainment.

REFERENCESABS. (2006). Adult Literacy and Life Skills Survey:

Summary Results. Australian Bureau of Statistics.Anderson, J. R. (1982). Acquisition of cognitive

skill. Psychol. Rev. 89, 369–406. doi: 10.1037/0033-295X.89.4.369

Arlington, K. (2012, February 20). Push to multiplypositive points of maths in minds of students. TheAustralian, p. 13.

Biggs, J. (1999). What the student does: teachingfor enhanced learning. Higher Educ. Res. Dev. 18,57–75. doi: 10.1080/0729436990180105

Butterworth, B. (2006). “Mathematical expertise”in Cambridge Handbook of Expertise and ExpertPerformance, ed K. A. Ericsson (Cambridge:Cambridge University Press), 553–568. doi:10.1017/CBO9780511816796.032

Clark, C. A. C., Nelson, J. M., Garza, J., Sheffield,T. D., Wiebe, S. A., and Espy, K. A. (2014).Gaining control: changing relations between exec-utive control and processing speed and their rel-evance for mathematics achievement over courseof the preschool period. Front. Psychol. 5:107. doi:10.3389/fpsyg.2014.00107

Ericsson, K. A. (1996). The Road to Excellence: TheAcquisition of Expert Performance in the Artsand Sciences, Sports, and Games. Mahwah, NJ:Erlbaum.


Furner, J. M., and Berman, B. T. (2003). Math anxiety:overcoming a major obstacle to the improvementof student math performance. Child. Educ. 79, 1–6.doi: 10.1080/00094056.2003.10522220

Griffin, S. (2009). Learning sequences in the acqui-sition of mathematical knowledge: using cog-nitive developmental theory to inform curricu-lum design for Pre-K–6 mathematics education.Mind Brain Educ. 3, 96–107. doi: 10.1111/j.1751-228X.2009.01060.x

Haase, V. G., Júlio-Costa, A., Lopes-Silva, J. B.,Starling-Alves, I., Antunes, A. M., Pinheiro-Chagas, P. et al. (2014). Contributions from spe-cific and general factors to unique deficits: twocases of mathematics learning difficulties. Front.Psychol. 5:102. doi: 10.3389/fpsyg.2014.00102

Healy, G. (2010, September 1). Survey finds graduates’mathematics doesn’t add up. The Australian, p.24.

Lei, P.-W., Wu, Q., DiPerna, J. C., and Morgan, P.L. (2009). Developing short forms of the EARLInumeracy measures comparison of item selectionmethods. Educ. Psychol. Meas. 69, 825–842. Doi:10.1177/0013164409332215

Logan, G. D. (1988). Toward an instance theory ofautomatization. Psychol. Rev. 95, 492–527. doi:10.1037/0033-295X.95.4.492

Main, S., and O’Rourke, J. (2011). New directions fortraditional lessons: can handheld game consolesenhance mental mathematics skills? Aust. J. Teach.Educ. 36, 42–55.

Maslen, G. (2012 May 15). Chief seeks new spark inscience. The Age, p. 13.

OECD (2013), PISA 2012 Results: What StudentsKnow and Can Do – Student Performance inMathematics, Reading and Science, Vol. I. Paris:OECD Publishing.

Neumann, M. M., Hood, M., Ford, R. M., andNeumann, D. L. (2013). Letter and numeral iden-tification: their relationship with early literacy andnumeracy skills. Eur. Early Child. Educ. Res. J. 21,489–501. doi: 10.1080/1350293X.2013.845438

Pellegrino, J. W., and Goldman, S. R. (1987).Information processing and elementary math-ematics. J. Learn. Disabil. 20, 23–32. doi:10.1177/002221948702000105

Ramsden, P. (1984). “The context of learning inacademic departments,” in The Experience OfLearning, eds F. Marton, D. Hounsell, and N.Entwistle (Edinburgh: Scottish Academic Press),198–216.

Rosas, R., Nussbaum, M., Cumsille, P., Marianov,V., Correa, M., Flores, P., et al. (2003). BeyondNintendo: design and assessment of educationalvideo games for first and second grade students.Comput. Educ. 40, 71–94. doi: 10.1016/S0360-1315(02)00099-4

Slattery, L. (2010 March 10). Equation for mathswarns of disaster. The Australian, p. 21.

Speelman, C. P. (2013). “A test of a computer gamedesigned to facilitate the acquisition of arithmeticskills,” in Enhancing Human Performance, ed C.P. Speelman (Cambridge: Scholars Publishing),1–21.

Speelman, C. P., and Kirsner, K. (2005). Beyondthe Learning Curve: The Construction ofMind. Oxford: Oxford University Press. doi:10.1093/acprof:oso/9780198570417.001.0001

Swan, P. (2004). “I hate mathematics,” in MAVAnnual Conference, Monash University. Availableonline at: http://www.mav.vic.edu.au/pd/confs/2004/ (Accessed September 12, 2012).

Towne, T. J., Ericsson, K. A., and Sumner, A.M. (2014). Uncovering mechanisms in videogame research: suggestions from the expert-performance approach. Front. Psychol. 5:161. doi:10.3389/fpsyg.2014.00161

van Kraayenoord, C. E., and Elkins, J. (2004).Learning difficulties in numeracy inAustralia. J. Learn. Dis. 37. 32–41. doi:10.1177/00222194040370010401

Whitburn, J. (1996). Contrasting approaches to theacquisition of mathematical skills: Japan andEngland. Oxford Rev. Educ. 22, 415–434. doi:10.1080/0305498960220403

Yelland, N., and Kildery, A. (2010). Becomingnumerate with information and commu-nications technologies in the twenty-firstcentury. Int. J. Early Years Educ. 18, 91–106.doi: 10.1080/09669760.2010.494426


Received: 15 March 2014; accepted: 25 May 2014;published online: 12 June 2014.Citation: Speelman CP (2014) The acquisition of exper-tise in the classroom: are current models of educationappropriate? Front. Psychol. 5:580. doi: 10.3389/fpsyg.2014.00580This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Speelman. This is an open-accessarticle distributed under the terms of the CreativeCommons Attribution License (CC BY). The use, dis-tribution or reproduction in other forums is permitted,provided the original author(s) or licensor are creditedand that the original publication in this journal is cited,in accordance with accepted academic practice. No use,distribution or reproduction is permitted which does notcomply with these terms.


http://www.mav.vic.edu.au/pd/confs/2004/

http://www.mav.vic.edu.au/pd/confs/2004/











OPINION ARTICLEpublished: 22 May 2014

doi: 10.3389/fpsyg.2014.00470

Designing a “better” brain: insights from experts andsavantsFernand Gobet1*, Allan Snyder 2, Terry Bossomaier 3 and Mike Harré4

1 Department of Psychological Sciences, University of Liverpool, Liverpool, UK2 Centre for the Mind, School of Medicine, University of Sydney, Sydney, NSW, Australia3 Centre for Research in Complex Systems, Faculty of Business, Charles Sturt University, Bathurst, NSW, Australia4 Complex Systems Research Group, Faculty of Engineering and IT, University of Sydney, Sydney, NSW, Australia*Correspondence: [email protected]



Keywords: autism, brain, cognitive illusion, creativity, Einstellung effect, expertise, perception, savants

If we naively, simply scale up the brain—more of the same—we would not necessar-ily have a brain that is qualitatively better.Taking the analogy from visual ecology—the eagle—it could just see better at greaterheights (Snyder and Miller, 1978)—moreof the same!

Instead, let us ask how to design a brainthat is qualitatively better in some sense,say at being creative. One bottleneck tocreativity is our inability to see things ina new light, free of prior interpretations.Once a pattern is identified and labeled,it is difficult to identify different patternsand find alternative solutions. A classicexample is the duck/rabbit illusion, wherewe can see either a duck or a rabbit butnot be simultaneously aware of both. Weare blinded by our mindsets, by our exper-tise! This presumably is a consequenceof our hypothesis-driven, cognitive makeup. Hypotheses are mental templates thatencapsulate expectations of the world asderived from past experience—a brilliantstrategy for coping rapidly with partialinformation in a dynamic but famil-iar environment (Snyder et al., 2004;Bossomaier et al., 2009). [Although we usethe human brain as an example, there isconsiderably evidence that the concept-dominated architecture is characteristic ofthe mammalian brain, and possibly alsoin birds, in particular corvids (e.g., Emeryand Clayton, 2004)].

Our brain can recall a seemingly unlim-ited number of meaningful patterns andlabels, but often not the attributes thatcompose them. In contrast, some autisticsavants can recall a seemingly unlimitedamount of details, without any attemptto impose meaning (e.g., Wiltshire, 1989).

But, savants are less prone to cognitiveillusions (Bogdashina, 2003) and that factgives us a clue for a “better” brain—a brain that is hypothesis driven, but isresilient to cognitive illusions, a brain thatcan in addition see the world with directperception and thus open to alternativeinterpretations (Snyder, 2009).

A better brain—natural or artificial—would have tremendous implications forour world. Most importantly, it wouldboost creativity, both in art and in science.Better decisions would be made in poli-tics and business, and long-standing issuessuch as pollution and hunger would bemore likely to be solved. How do we goabout to developing such a brain?

A possible strategy is to start from thethings humans are doing so well as a resultof our adaptation to the environment dueto evolutionary pressure, and then to con-sider why the products of this adaptationsometimes are associated with penalties.Finding a way to fix these instances ofpenalties in humans leads to insights thatcan be further applied to the design ofbetter brains.

As a first example, consider visual per-ception. Overall, our eye and our visualcortex, honed by millions of years of evo-lution, work very well. However, humanvisual perception can be fooled surpris-ingly easily by perceptual and cognitiveillusions. We can see things that do notexist (e.g., Escher’s impossible figures), donot see things that exist (e.g., readingwords rather than seeing black and whitepixels), see two objects from the samedrawing (e.g., the duck-rabbit illusion),and grossly misjudge the dimensions ofobjects (e.g., the Ebbinghaus and Ponzo

illusions). How can we avoid these illu-sions? What would this tell us about braindesign?

As a second example, consider exper-tise. Experts can show extreme adapta-tions to their environment and are capa-ble of great achievements (Ericsson et al.,2006; Didierjean and Gobet, 2008). Tennisplayers can return services with ballshit at more than 200 km/h, and theycan even counter-attack when doing so.Mnemonists can memorize more than 100digits read at 1 s each. Some chess playerscan play more than 30 games simultane-ously, without seeing the board and thepieces. But expertise sometimes fails. Intheir study on the Einstellung effect inchess, Bilalic et al. (2008a,b, 2010) showedthat even experts can fail to find an opti-mal solution when a common solutioncomes first to their mind. The effect issurprisingly powerful: compared to con-trol positions, players lose about 1 stan-dard deviation in skill when showingthe Einstellung effect. This illustrates thestrength of the schemas we hold in long-term memory and the power that ourpreconceptions have on our mind.

What are the solutions to percep-tual/cognitive illusions, such as thosedisplayed in visual illusions and in theEinstellung effect? Intuitively, there aretwo main approaches. The first one is touse more knowledge (i.e., scaling up) andcould be called “quantitative scaling.” Itcould be summarized by the phrase “moreof the same.” This is the standard approachin computer science and artificialintelligence.

Quantitative scaling has had sometremendous successes in artificial

www.frontiersin.org May 2014 | Volume 5 | Article 470 | 173












Gobet et al. Designing a “better” brain

intelligence, such as the victory of DeepBlue against chess world championKasparov in 1997 (Campbell et al., 2002).But increase of knowledge does not alwaysprovide a solution; for example, manyvisual illusions persist in spite of ourknowledge that they exist. Indeed, quanti-tative scaling has also met with problemsin artificial intelligence. Most notably, ithas failed in many attempts, such as play-ing Go at master level (Gobet et al., 2004;Hsu, 2007), providing computers withcommon sense (McCarthy, 2007; Sarbo,2007), and in general developing genuine,domain-general intelligence and creativity(Bridewell and Langley, 2010; Jennings,2010).

A second approach could be called“qualitative scaling.” In essence, this leadsto the definition and use of new concep-tual spaces. To get to these new spaces, weneed to break apart the existing conceptualstructures, whereas in quantitative scal-ing we build increasingly elaborate con-cepts on top of what we have already.A classic example is Einstein’s conjec-ture that the speed of light was constantin all inertial reference frames, leadingto conclusions very different to those ofNewtonian physics. Serialism, cubism orsimply Dadaism itself illustrate this notionin the arts.

We know that individuals with autismare less susceptible to some illusions(Walter et al., 2009; Mitchell et al., 2010),and that they process information in a lessholistic way (Nakahachi et al., 2008), forexample paying more attention to details,which can result in savant skills such asextreme memory for detail and speed incounting objects (numerosity) (Souliereset al., 2010). We also know that somesavant skills can be artificially induced innormal individuals, for example by low-frequency repetitive transcranial magneticstimulation (Snyder et al., 2006; Boggioet al., 2009; Snyder, 2009) or transcra-nial direct current stimulation (Chi andSnyder, 2012).

Our key argument is that our nor-mal cognition, while very efficient, tendsto develop cognitive mind sets. Breakingthese mind sets can help explore newconceptual spaces, and thus be more cre-ative. Rather than reorganizing knowledgein some superficial way, we propose tworadical approaches. The first is to inhibit

some concepts or a class of concepts (e.g.,a group of concepts that are stronglyconnected or the most likely concepts ina given situation). The second is to avoidconcepts altogether (or at least as much aspossible), by using raw perception instead,thus imitating the autistic mind. Our pro-posal is thus that creativity can be boostedby decreasing conceptual processing andincreasing the role of low-level perceptualprocessing.

Our proposal raises intriguing issues.What is the link between concepts and rawperception? Are they two discrete states,or are they part of a more graded space?How can “raw perception” mitigate oreven eliminate Einstellung-like effects? Israw perception enough? Autism seems tooffer a counter-example, where the lackof use of concepts leads to serious intel-lectual and social impairments. But is itnecessarily so?

The history of human thought pro-vides us with creativity examples wherethe inhibiting-concept strategy was used(possibly unconsciously) and worked.When 2005 Nobel Laureates Marshalland Warren (1984) correctly proposedthat stomach ulcers were caused by heli-cobacter pylori rather than by excessacid, they had to jettison a whole raftof concepts. There are also examples ofthis in AI research—for instance com-puter chess and checkers, where newprofound insights were gained from bruteforce search, without using sophisticatedconcepts. In the latter case, the literal per-ception used by computers, which is oftenderided, turns out to be a strength. Thecost of deleting concepts might also bestudied. If the concepts that are inhibitedare infrequent and are not the buildingblocks of a large number of other concepts,the benefits might outweigh the costs. Butwhat is the threshold? Another solutionwith artificial systems is parallelism. Apossibility would be that the original con-ceptual base stays online while conceptsare inhibited in a copy of the original basethat operates offline.

The examples in this article focus onfinding novel creative solutions withoutbeing bound by prior knowledge. It is anopen question as to exactly what a betterbrain would be: more creative, more effi-cient, more rational, more adaptive, morealtruistic? Answering this question is a

huge challenge in itself, and we should bealert to our own mindset in defining thespace of possible answers.

The implications for the study ofhuman cognition and the psychologyexpertise in particular are profound. A bet-ter brain would shed considerable but alsocruel light on the limits of human cog-nition and expertise. We already had apreview of this with developments in com-puter chess. No players, even world cham-pion Magnus Carlsen, can compete withcomputers nowadays. Computers some-times find moves that are considered byhumans as highly creative, although someof these moves are just beyond humandiscovery. In addition, computers haveled to re-evaluations of large aspects ofthe game, in particular openings andendgames, which humans had researchedfor centuries. Be ready to be surprised withbetter brains!

Irrespective of possible benefits for sci-ence and technology, including artificialintelligence, our proposal for a better brainraises important questions for the natureof the human mind. Are qualitative scalingand quantitative scaling really in opposi-tion? Is it adaptive to inhibit some con-cepts “just” to be creative? Why has such asystem not evolved? Is it because, now, wehave the luxury to be creative—courtesy ofcultural evolution?

ACKNOWLEDGMENTSFernand Gobet was supported by aResearch Fellowship from the UKEconomic and Social Research Council.

REFERENCESBilalic, M., McLeod, P., and Gobet, F. (2008a).

Inflexibility of experts–reality or myth?Quantifying the Einstellung effect in chessmasters. Cogn. Psychol. 56, 73–102. doi: 10.1016/j.cogpsych.2007.02.001

Bilalic, M., McLeod, P., and Gobet, F. (2008b). Whygood thoughts block better ones: the mecha-nism of the pernicious Einstellung (set) effect.Cognition 108, 652–661. doi: 10.1016/j.cognition.2008.05.005

Bilalic, M., McLeod, P., and Gobet, F. (2010). Themechanism of the einstellung (set) effect: a perva-sive source of cognitive bias. Curr. Dir. Psychol. Sci.19, 111–115. doi: 10.1177/0963721410363571

Bogdashina, O. (2003). Sensory Perceptual Issues inAutism and Asperger Syndrome. London: Kingsley.

Boggio, P. S., Fregni, F., Valasek, C., Ellwood, S., Chi,R., Gallate, J., et al. (2009). Temporal lobe corti-cal electrical stimulation during the encoding andretrieval phase reduces false memories. PLoS ONE4:e4959. doi: 10.1371/journal.pone.0004959

Frontiers in Psychology | Cognition May 2014 | Volume 5 | Article 470 | 174




Gobet et al. Designing a “better” brain

Bossomaier, T., Harre, M., Knittel, A., and Snyder, A.(2009). A semantic network approach to the cre-ativity quotient (CQ). Creat. Res. J. 21, 64–71. doi:10.1080/10400410802633517

Bridewell, W., and Langley, P. (2010). Two kinds ofknowledge in scientific discovery. Top. Cogn. Sci. 2,36–52. doi: 10.1111/j.1756-8765.2009.01050.x

Campbell, M., Hoane, A. J., and Hsu, F. H.(2002). Deep blue. Artif. Intell. 134, 57–83. doi:10.1016/S0004-3702(01)00129-1

Chi, R., and Snyder, A. (2012). Brain stimulationenables the solution of an inherently difficult prob-lem. Neurosci. Letters 515, 121–124. doi: 10.1016/j.neulet.2012.03.012

Didierjean, A., and Gobet, F. (2008). Sherlock Holmes- an expert’s view of expertise. Br. J. Psychol. 99,109–125. doi: 10.1348/000712607X224469

Emery, N. J., and Clayton, N. S. (2004). The mental-ity of crows: convergent evolution of Intelligencein corvids and apes. Science 306, 1903–1907. doi:10.1126/science.1098410

Ericsson, K. A., Charness, N., Feltovich, P. J.,and Hoffman, R. R. (2006). The CambridgeHandbook of Expertise. New York, NY: CUP. doi:10.1017/CBO9780511816796

Gobet, F., De Voogt, A. J., and Retschitzki, J. (2004).Moves in Mind. Hove: Psychology Press.

Hsu, F. H. (2007). Cracking go. IEEE Spectr. 44, 50–55.doi: 10.1109/MSPEC.2007.4337666

Jennings, K. E. (2010). Developing creativity: artificialbarriers in artificial intelligence. Minds Mach. 20,489–501. doi: 10.1007/s11023-010-9206-y

Marshall, B. J., and Warren, J. R. (1984). Unidentifiedcurved bacilli in the stomach patients with gastri-tis and peptic ulceration. Lancet 1, 1311–1315. doi:10.1016/S0140-6736(84)91816-6

McCarthy, J. (2007). From here to human-level AI. Artif. Intell. 171, 1174–1182. doi:10.1016/j.artint.2007.10.009

Mitchell, P., Mottron, L., Soulieres, I., and Ropar, D.(2010). Susceptibility to the Shepard illusion inparticipants with autism: reduced top-down influ-ences within perception? Autism Res. 3, 113–119.doi: 10.1002/aur.130

Nakahachi, T., Yamashita, K., Iwase, M., Ishigami, W.,Tanaka, C., Toyonaga, K., et al. (2008). Disturbedholistic processing in autism spectrum disordersverified by two cognitive tasks requiring percep-tion of complex visual stimuli. Psychiatry Res. 159,330–338. doi: 10.1016/j.psychres.2005.08.028

Sarbo, J. J. (2007). “On the logic underlyingcommon sense,” in ICEIS 2007: Proceedings ofthe Ninth International Conference on EnterpriseInformation Systems - Information Systems Analysisand Specification, eds J. Cardoso, J. Cordeiro, andJ. Filipe (Funchal), 395–400.

Snyder, A. (2009). Explaining and inducing savantskills: privileged access to lower level, less-processed information. Philos. Trans. R. Soc.B. Biol. Sci. 364, 1399–1405. doi: 10.1098/rstb.2008.0290

Snyder, A., Bawamali, H., Hawker, T., and Mitchell, D.J. (2006). Savant-like numerosity skills revealed innormal people by magnetic pulses. Perception 35,837–845. doi: 10.1068/p5539

Snyder, A., Mitchell, J., Bossomaier, T., and Pallier, G.(2004). The creativity quotient: an objective scor-ing of ideational fluency. Creat. Res. J. 16, 415–420.doi: 10.1080/10400410409534552

Snyder, A. W., and Miller, W. H. (1978). Telephoto lenssystem of falconiform eyes. Nature 275, 127–129.doi: 10.1038/275127a0

Soulieres, I., Hubert, B., Rouleau, N., Gagnon,L., Tremblay, P., Seron, X., et al. (2010).Superior estimation abilities in two autis-tic spectrum children. Cogn. Neuropsychol.27, 261–276. doi: 10.1080/02643294.2010.519228

Walter, E., Dassonville, P., and Bochsler, T. M.(2009). A specific autistic trait that modulatesvisuospatial illusion susceptibility. J. Autism Dev.Disord. 39, 339–349. doi: 10.1007/s10803-008-0630-2

Wiltshire, S. (1989). Cities. London: Dent.


Received: 03 March 2014; accepted: 01 May 2014;published online: 22 May 2014.Citation: Gobet F, Snyder A, Bossomaier T and Harré M(2014) Designing a “better” brain: insights from expertsand savants. Front. Psychol. 5:470. doi: 10.3389/fpsyg.2014.00470This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Gobet, Snyder, Bossomaier andHarré. This is an open-access article distributed underthe terms of the Creative Commons Attribution License(CC BY). The use, distribution or reproduction in otherforums is permitted, provided the original author(s) orlicensor are credited and that the original publicationin this journal is cited, in accordance with acceptedacademic practice. No use, distribution or reproduc-tion is permitted which does not comply with theseterms.

www.frontiersin.org May 2014 | Volume 5 | Article 470 | 175












doi: 10.3389/fpsyg.2014.00573

2011 space odyssey: spatialization as a mechanism to codeorder allows a close encounter between memory expertiseand classic immediate memory studiesAlessandro Guida1* and Magali Lavielle-Guida2

1 Psychology Department, CRPCC, Université Rennes 2, Rennes, France2 Cabinet de Psychologie et d’Orthophonie, St Malo, France*Correspondence: [email protected]; [email protected]



Keywords: spatialization, immediate memory, expertise, long-term working memory, retrieval structures

In 2011 van Dijk and Fias with an inno-vative working memory paradigm showedfor the first time that words to-be-remembered, presented sequentially at thecenter of a screen acquired a new spatialdimension: the first words of the sequenceacquired a left spatial value while the lastwords acquired a right spatial value. In thisarticle, we argue that this spatializationwhich putatively underpins how order iscoded in immediate memory1 allows brid-ging the domain of memory expertise withclassic immediate memory studies.

After briefly reviewing the mecha-nisms for coding order in immediatememory and the recent studies pointingtoward spatialization as an explanatorymechanism, we will pinpoint similarmechanisms that are known to exist inmemory expertise, particularly in themethod of loci. We will terminate by ana-lyzing what these similarities can tell usabout expertise.

HOW ORDER IS CODED?Surprisingly, this very fundamental ques-tion has not yet received a definitiveanswer. If one tries to naively think abouta way order could be coded, generallythe first idea that comes is chaining:items in a list to-be-remembered are justchained together by our cognitive system.And indeed, for more than four decades,this has been the most prominent ideaamong researchers (e.g., Wickelgren, 1965;Jordan, 1986; Lewandowsky and Murdock,1989). This idea beyond being simple and

1 Immediate memory is an umbrella term for workingmemory and short-term memory.

intuitive, is also ancient since it rootsback at least to Ebbinghaus (1885/2010).However, in the last two decades chainingmodels have lost ground, mostly becauseof experimental results. In immediatememory, error patterns (i.e., transposi-tion and protrusion errors, Estes, 1991;Henson, 1996, 1999) and the distanceeffect (e.g., Hacker, 1980; Marshuetz et al.,2000) have been difficult to explain withthe chaining concept.

POSITIONAL TAGGINGNowadays prominent models are of a posi-tional kind (e.g., Anderson and Matessa,1997; Burgess and Hitch, 1999; Brownet al., 2000, 2007; O’Reilly and Soto,2001; Lewandowsky and Farrell, 2008a;Oberauer and Lewandowsky, 2011).Based on various studies (e.g., Dale,1987; Poirier and Saint-Aubin, 1996;Mulligan, 1999; Engelkamp and Dehn,2000; Henson et al., 2003), these mod-els assume that item information andorder information are coded and rep-resented separately (for a review, seeMarshuetz, 2005). Order is putativelycoded through positional coding mech-anisms, where a positional marker (ortag)–a context–is associated to each item.These contexts or positional markerscan be temporal or not (Lewandowskyand Farrell, 2008b), but several studiesseem to run against temporal markers(e.g., Lewandowsky and Brown, 2004,2005; Lewandowsky et al., 2006), whichfavors non-temporal ones. Nonethelessif temporal tags are by definition well-known, the nature of non-temporal tagsremains unknown (Lewandowsky and

Farrell, 2008b)2. It could be an externalcontext such as the environment or/andan internal context such as the innerstates of the mind associated with eachitems.

WHAT DOES VAN DIJCK AND FIAS(2011) STUDY CHANGE CONCERNINGORDER CODING?In 2011 van Dijck and Fias pro-posed an alternative explanation of theSNARC (Spatial-Numerical Associationof Response Codes) effect. This effectwas first popularized by Dehaene et al.(1993). They used a classic parity judg-ment task where participants had todecide if a number was odd or even.However, the left-/right-hand key assign-ment was varied: the answer “even” (asthe answer “odd”) was assigned for halfof the trials to one hand and for the otherhalf to the other hand. Results showeda SNARC effect, that is, small numberstriggered faster responses when partic-ipants answered with the left hand andlarge numbers triggered faster responseswhen participants answered with the righthand. According to Dehaene et al. (1993),the effect was due to the representation

2 Lewandowsky and Farrell (2008b) wrote: “The useof context markers does, however, entail a cost: As inmany other models (e.g., SEM; Henson, 1998), thestructure of the markers across positions is assumedrather than explained by the model. That is, althoughit is entirely plausible to postulate that the contextsof adjacent items are more similar to each other thanthe contexts of items separated by intervening events,the precise form of their similarity relationship is notderived from the model’s architecture. Are there anycandidate mechanisms on the horizon that might per-mit a more principled derivation of context markers?.”










mailto:[email protected]; [email protected]



Guida and Lavielle-Guida Spatialization and immediate memory

numbers have in (semantic) long-termmemory (LTM), that of a mental line,which in western cultures increases fromleft to right (e.g., Dehaene et al., 1993;Göbel et al., 2011).

This LTM conception of the SNARCwas disputed by van Dijck and Fias (2011)using a new paradigm. They proposedthat the SNARC effect depended on theorganization numbers assume in workingmemory. In the study, participants werepresented five random numbers (rangingfrom 1 to 10) to-be-remembered in cor-rect order. Numbers were displayed at thecenter of a screen. After the presentationphase, numbers ranging from 1 to 10 weredisplayed randomly at the center of screen.When a number to-be-remembered wasdisplayed, participants had to execute aparity judgment task. As in Dehaene et al.(1993), the left-/right-hand key assign-ment was varied. But instead of the

usual SNARC effect, results showed aSpatial-Positional Association of ResponseCodes (SPoARC) effect, that is, left handresponses were faster with numbers pre-sented in the first positions of the to-be-remembered numbers (instead of smallnumbers in the SNARC effect) and righthand responses were faster with numberspresented in the last positions (instead ofbig numbers).

A NEW POSITIONAL TAGGINGMECHANISM: SPATIALIZATIONThis result and others (i.e., van Dijck et al.,2013; Guida, under review) suggest thatthe initial words of a sequence have aleft spatial value while the last wordsof the same sequence have a right spa-tial value. Apparently individuals tendto create a spatial mental line based onthe order items enter immediate mem-ory (Example 1, Figure 1). This is highly

compatible with the idea that in verbalimmediate memory, items order is codedspatially, through spatialization. Giventhe fuzzy nature of non-temporal tags,this discovery could allow specifying theway items order is coded in immediatememory.

WHAT HAS SPATIALIZATION GOT TODO WITH MEMORY EXPERTISE?Since the very first (internal) mnemonic(Yates, 1966; Worthen and Hunt, 2011)which is thought to be the loci methodproposed by Simonides of Ceos (556 BC–448 BC) and reported by Marcus TulliusCicero in De Oratore, visuo-spatial pro-cesses have played a central role to enhancememory for verbal material. Concerningthe loci method, Simonides of Ceos pro-posed to visualize a familiar route or asequence of familiar locations (like roomsin one’s own house) and use them to

FIGURE 1 | Schematic representation of retrieval structures through two

examples. The upper part of the figure offers a generic and abstractrepresentation of retrieval structures, from Ericsson and Kintsch (1995). Thefirst example is taken from the spatial positional mental line and adapted

from Guida (under review), it represent the encoding of three letters via threespatial positional tags. The second example is from the method of loci, andrepresents the encoding of three words via known locations used as retrievalcues.






mentally store a list of words (Example2, Figure 1), before a speech for example.Then during the speech, one would takea mental tour and retrieve each wordvia each familiar location (e.g., kitchen).Greek orators (Yates, 1966; Worthen andHunt, 2011) soon became experts of themethod of loci.

THE METHOD OF LOCI: EXPERTISETHROUGH SPATIALIZATIONOf interest here, is the fact that the locimethod necessitates to spatialize the itemsto-be-remembered in various locations.Moreover the method is not just an ancientoddity, the method efficiency has beenconfirmed since and still is nowadays.Memory experts (i.e., mnemonists) use itand several memory world records havebeen set with it. For example Pridmore(2013) was the first man to break the 30-s barrier in the Speed Cards discipline,which necessitates to memorize the orderof a shuffled deck of cards. To do so, heused a system based on the method of loci,he spatialized groups of two cards the longof a familiar route.

IS THERE A THEORY OF EXPERTISETHAT SUPPORTS THE LOCI METHODPHENOMENOLOGY WHICH POINTSTOWARD SPATIALIZATION?Even if mnemonics and memory exper-tise are very ancient (certainly due to oraltradition, see Rubin, 1997; Ong, 2012),grounded cognitive theories describingthem are recent. It could be argued thatthe first complete theoretical contribu-tion on mnemonic expertise (but seethe Chunking theory, Chase and Simon,1973) was Chase and Ericsson’s (1981)Skilled memory theory, which was to becompleted with the Long-term workingmemory (LT-WM) theory (Ericsson andKintsch, 1995).

LONG-TERM WORKING MEMORY ANDRETRIEVAL STRUCTURESIn order to explain memory expertise,Chase and Ericsson (1981) proposed threeprinciples: the significant encoding, thestructured retrieval and the principle ofacceleration. The first principle proposesthat in order to swiftly and reliablystore items in LTM, information needto be transformed into meaningful units.Of interest here is the second principle

which states that to increase mnemonicperformances, hierarchical spatial cogni-tive structures, named retrieval struc-tures (for a discussion, see Ericsson andKintsch, 2000; Gobet, 2000a,b) can beused to encode and retrieve items fromLTM. These structures constitute an inter-nal artificial context to which items arelinked to. In the loci method, it isdone via the visuo-spatial knowledge of asequence of familiar locations. Each loca-tion is a retrieval cue, and all the cuestogether constitute a retrieval structure(Figure 1). The skilled memory theory wasfirst proposed to account for the per-formances of experts capable to increasetheir digit span above 80. The LT-WM(Ericsson and Kintsch, 1995) was a gen-eralization of this theory to all activi-ties and to all individuals, experts andnovices.

WHAT DOES SPATIALIZATION AS ALINK BETWEEN CLASSIC IMMEDIATEMEMORY STUDIES AND MEMORYEXPERTISE BRING ASPSYCHOLOGICAL PERSPECTIVES ONEXPERTISE?Notwithstanding Ericsson and Kintsch’s(1995) generalization, the LT-WM the-ory remains underused in the classicdomain of verbal immediate memory(but, see Guida et al., 2009, 2013). Asstated by Ericsson and Kintsch (1995, p.217) concerning the Skilled memory the-ory (but the same can be said for LT-WM), even if this theoretical construct islargely accepted as accounting for experts,“several investigators (e.g., Schneider andDetweiler, 1987; Carpenter and Just, 1989;Baddeley, 1990) have voiced doubts aboutits generalizability.” Retrieval structuresare often dismissed because considered tooartificial or idiosyncrasies to be reservedto experts. Thank to van Dijck and Fias’s(2011) study, this could change.

RETRIEVAL STRUCTURE ASSPATIALIZATION: A GENUINE ANDUNIVERSAL PROCESSAs seen previously, van Dijck and col-leagues’ results (van Dijck and Fias, 2011;van Dijck et al., 2013; see also Guida,under review) clearly point toward the ideathat in all-comers, spatial processes arealso at stake in verbal immediate mem-ory. When comparing retrieval structures

such as in the method of loci and spatialpositional tags, the similarities are strik-ing (Figure 1). In both cases, a virtualspatial construct, used as a context, is asso-ciated to the incoming information. Andthe context can later be used to retrievethe items. Even if the mental line (Dehaeneet al., 1993; van Dijck and Fias, 2011) usedby all-comers is much simpler and lessersophisticated, compared to mnemonistsusing the method of loci, spatializationseems the same underpinning process. Ifthis standpoint is adopted, then it becomesmore explicit why the loci method is soancient and efficient: because experts’ spa-tialization via retrieval structures roots onbasic processes that all individuals can use.Ipso facto, retrieval structures stop beingidiosyncrasies to be reserved to experts.

The link between both kinds of spa-tialization becomes even more tangiblewhen considering that the spatial men-tal line could also be due to our exper-tise, in this case in mastering the writingsystem. In fact the orientation and direc-tion of our mental line varies accordingto reading/writing habits (e.g., Dehaeneet al., 1993; Shaki et al., 2009; Göbel et al.,2011; for the influence of reading habits onvisuo-spatial processes, e.g., see Maass andRusso, 2003; Dobel et al., 2007). Therefore,it is very plausible that our reading andwriting habits foster our spatial mentalline.

When considering the privileged linkbetween space and memory, it is also inter-esting to conclude taking a brief glance toanthropology, which shows that this linkseems to be far more ancient than ourreading habits and already present in non-literate societies. In fact myths around theworld have often been linked to specificlocations. This “myth spatialization” canbe found in the Tobriand culture fromPapua New Guinea for example, or inthe Australian aborigines famous song-lines (Chatwin, 1987) or even in Zunis’legends from southwestern United States.In all these cases, “spatial location func-tions as a mnemonic device for the recallof a corpus of myth” (Harwood, 1976, p.783). Building on Harwood’s (1976) mythspatialization, the loci method can be con-sidered as a phylogenetic protraction of themyth spatialization, and the mental spatialline as an ontogenetic protraction of ourreading habits.





REFERENCESAnderson, J. R., and Matessa, M. (1997). A pro-

duction system theory of serial memory. Psychol.Rev. 104, 728–748. doi: 10.1037/0033-295X.104.4.728

Baddeley, A. D. (1990). Human Memory: Theory andPractice. Boston, MA: Allyn & Bacon.

Brown, G. D. A., Neath, I., and Chater, N. (2007).A temporal ratio model of memory. Psychol.Rev. 114, 539–576. doi: 10.1037/0033-295X.114.3.539

Brown, G. D. A., Preece, T., and Hulme, C.(2000).Oscillator-based memory for serial order.Psychol. Rev. 107, 127–181. doi: 10.1037/0033-295X.107.1.127

Burgess, N., and Hitch, G. J. (1999). Memory for serialorder a network model of the phonological loopand its timing. Psychol. Rev. 106, 551–581. doi:10.1037/0033-295X.106.3.551

Carpenter, P. A., and Just, M. A. (1989). “The roleof working memory in language comprehension,”in Complex Information Processing: The Impact ofHerbert A. Simon, eds D. Klahr and K. Kotovsky(Hillsdale, NJ: Erlbaum), 31–68.

Chase, W. G., and Ericsson, K. A. (1981). “Skilledmemory,” in Cognitive Skills and their Acquisition,ed J. R. Anderson (Hillsdale, NJ: LawrenceErlbaum Associates), 141–189.

Chase, W. G., and Simon, H. A. (1973). Perception inchess. Cogn. Psychol. 4, 55–81. doi: 10.1016/0010-0285(73)90004-2

Chatwin, B. (1987) The Songlines. London: JonathanCape Ltd.

Dale, R. H. (1987). Similarities between human andanimal spatial memory: item and order infor-mation. Anim. Learn. Behav. 15, 293–300. doi:10.3758/BF03205022

Dehaene, S., Bossini, S., and Giraux, P. (1993). Themental representation of parity and numericalmagnitude. J. Exp. Psychol. Gen. 122, 371–396. doi:10.1037/0096-3445.122.3.371

Dobel, C., Diesendruck, G., and Bolte, J. (2007).How writing system and age influence spatial rep-resentations of actions: a developmental, cross-linguistic study. Psychol. Sci. 18, 487–491. doi:10.1111/j.1467-9280.2007.01926.x

Ebbinghaus, H. (1885/2010). La Mémoire: Recherchesde Psychologie Expérimentale. Paris: L’Harmattan.

Engelkamp, J., and Dehn, D. M. (2000). Itemand order information in subject-performedtasks and experimenter-performed tasks. J. Exp.Psychol. Learn. Mem. Cogn. 26, 671–682. doi:10.1037/0278-7393.26.3.671

Ericsson, K. A., and Kintsch, W. (1995). Long-termworking memory. Psychol. Rev. 102, 211–245. doi:10.1037/0033-295X.102.2.211

Ericsson, K. A., and Kintsch, W. (2000). Shortcomingsof generic retrieval structures with slots of thetype of Gobet (1993) proposed and modeled.Br. J. Psychol. 91, 571–590. doi: 10.1348/000712600161998

Estes, W. K. (1991). “On types of item codingand sources of recall in short-term memory,” inRelating Theory and Data: In Honor of Bennet B.Murdock, eds W. E. Hockley and S. Lewandowsky(Hillsdale, NJ: Lawrence Erlbaum Associates Inc.),175–194.

Göbel, S. M., Shaki, S., and Fischer, M. H. (2011).The cultural number line: a review of cultural and

linguistic influences on the development of num-ber processing. J. Cross Cult. Psychol. 42, 543–565.doi: 10.1177/0022022111406251

Gobet, F. (2000a). Some shortcomings of long-termworking memory. Br. J. Psychol. 91, 551–570. doi:10.1348/000712600161989

Gobet, F. (2000b). Retrieval structures and schemata:a brief reply to Ericsson and Kintsch. Br. J.Psychol. 91, 591–594. doi: 10.1348/000712600162005

Guida, A., Gras, D., Noel, Y., Le Bohec, O.,Quaireau, C., and Nicolas, S. (2013). Theeffect of long-term working memory throughpersonalization applied to free recall: uncurb-ing the primacy effect enthusiasm. Mem.Cognit. 41, 571–587. doi: 10.3758/s13421-012-0284-3

Guida, A., Tardieu, H., and Nicolas, S. (2009).The personalisation method applied to aworking memory task: evidence of long-term working memory effects. Eur. J. Cogn.Psychol. 21, 862–896. doi: 10.1080/09541440802236369

Hacker, M. J. (1980). Speed and accuracy of recencyjudgments for events in short-term memory.J. Exp. Psychol. Hum. Learn. 6, 651–675. doi:10.1037/0278-7393.6.6.651

Harwood, F. (1976). Myth, memory, and the oraltradition: Cicero in the Trobriands. Am. Anthropol.78, 783–796. doi: 10.1525/aa.1976.78.4.02a00040

Henson, R. N. A. (1996). Short-Term Memory forSerial Order. Unpublished doctoral dissertation,University of Cambridge, Cambridge.

Henson, R. N. A. (1998). Short-term memory forserial order: the start-end model. Cogn. Psychol. 36,73–137. doi: 10.1006/cogp.1998.0685

Henson, R. N. A. (1999). Coding position in short-term memory. Int. J. Psychol. 34, 403–409. doi:10.1080/002075999399756

Henson, R. N. A., Hartley, T., Burgess, N., Hitch,G., and Flude, B. (2003). Selective interferencewith verbal short-term memory for serial orderinformation: a new paradigm and tests of atiming signal hypothesis. Q. J. Exp. Psychol.A 56, 1307–1334. doi: 10.1080/02724980244000747

Jordan, M. I. (1986). Serial Order: a ParallelDistributed Approach (ICS Report 8604). SanDiego: University of California, Institute forCognitive Science.

Lewandowsky, S., and Brown, G. D. A. (2004).Time does not cause forgetting in short-termserial recall. Psychon. Bull. Rev. 11, 771–790. doi:10.3758/BF03196705

Lewandowsky, S., and Brown, G. D. A. (2005). Serialrecall and presentation schedule: a micro-analysisof local distinctiveness. Memory 13, 283–292. doi:10.1080/09658210344000251

Lewandowsky, S., Brown, G. D. A., Wright, T.,and Nimmo, L. M. (2006). Timeless memory:Evidence against temporal distinctiveness mod-els of short-term memory for serial order. J.Mem. Lang. 54, 20–38. doi: 10.1016/j.jml.2005.08.004

Lewandowsky, S., and Farrell, S. (2008a).Phonological similarity in serial recall: con-straints on theories of memory. J. Mem. Lang. 58,429–448. doi: 10.1016/j.jml.2007.01.005

Lewandowsky, S., and Farrell, S. (2008b). “Short-termmemory: new data and a model,” in The Psychologyof Learning and Motivation, Vol. 49, ed B. H. Ross(London: Elsevier), 1–48.

Lewandowsky, S., and Murdock, B. B. Jr. (1989).Memory for serial order. Psychol. Rev. 96, 25–57.doi: 10.1037/0033-295X.96.1.25

Maass, A., and Russo, A. (2003). Directional biasin the mental representation of spatial events:nature or culture? Psychol. Sci. 14, 296–301. doi:10.1111/1467-9280.14421

Marshuetz, C. (2005). Order information inworking memory: an integrative review ofevidence from brain and behavior. Psychol.Bull. 131, 323–339. doi: 10.1037/0033-2909.131.3.323

Marshuetz, C., Smith, E. E., Jonides, J., DeGutis,J., and Chenevert, T. L. (2000). Order infor-mation in working memory: fMRI evidence forparietal and prefrontal mechanisms. J. Cogn.Neurosci. 12, 130–144. doi: 10.1162/08989290051137459

Mulligan, N. W. (1999). The effects of percep-tual interference at encoding on organizationand order: Investigating the roles of item-specific and relational information. J. Exp. Psychol.Learn. Mem. Cogn. 25, 54–69. doi: 10.1037/0278-7393.25.1.54

Oberauer, K., and Lewandowsky, S. (2011). Modelingworking memory: a computational implementa-tion of the time-based resource-sharing theory.Psychon. Bull. Rev. 18, 10–45. doi: 10.3758/s13423-010-0020-6

Ong, W. J. (2012). Orality and Literacy: TheTechnologizing of the Word. London: Routledge.

O’Reilly, R. C., and Soto, R. (2001). “A model ofthe phonological loop: generalization and bind-ing,” in Advances in Neural Information ProcessingSystems, eds T. G. Dietterich, S. Becker, andZ. Ghahramani (Cambridge, MA: MIT Press),83–90.

Poirier, M., and Saint-Aubin, J. (1996). Immediateserial recall, word frequency, item identity anditem position. Can. J. Exp. Psychol. 50, 408–412.doi: 10.1037/1196-1961.50.4.408

Pridmore, B. (2013). How to be Clever. Availableonline at: http://www.lulu.com/

Rubin, D. C. (1997). Memory in Oral Traditions:The Cognitive Psychology of Epic, Ballads, andCounting-out Rhymes. New York, NY: OxfordUniversity Press.

Schneider, W., and Detweiler, M. (1987). “A connec-tionist/control architecture for working memory,”in The Psychology of Learning and Motivation, edG. H. Bower (New York, NY: Academie Press),54–119.

Shaki, S., Fischer, M. H., and Petrusic, W. M. (2009).Reading habits for both words and numbers con-tribute to the SNARC effect. Psychon. Bull. Rev. 16,328–331. doi: 10.3758/PBR.16.2.328

van Dijck, J.-P., Abrahamse, E. L., Majerus, S.,and Fias, W. (2013). Spatial attention interactswith serial order retrieval in verbal workingmemory. Psychol. Sci. 24, 1854–1859. doi:10.1177/0956797613479610

van Dijck, J. P., and Fias, W. (2011). A work-ing memory account for spatial-numericalassociations. Cognition 119, 114–119. doi:10.1016/j.cognition.2010.12.013


http://www.lulu.com/





Wickelgren, W. A. (1965). Short-term memory forphonemically similar lists. Am. J. Psychol. 78,567–574. doi: 10.2307/1420917

Worthen, J. B., and Hunt, R. R. (2011). Mnemonology:Mnemonics for the 21st Century. Hove: PsychologyPress.

Yates, F. A. (1966). The Art of Memory. Chicago:University of Chicago Press.

Conflict of Interest Statement: The authors declarethat the research was conducted in the absence

of any commercial or financial relationshipsthat could be construed as a potential conflict ofinterest.

Received: 02 May 2014; accepted: 23 May 2014;published online: 10 June 2014.Citation: Guida A and Lavielle-Guida M (2014) 2011space odyssey: spatialization as a mechanism to codeorder allows a close encounter between memory exper-tise and classic immediate memory studies. Front.Psychol. 5:573. doi: 10.3389/fpsyg.2014.00573

This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Guida and Lavielle-Guida. This isan open-access article distributed under the terms of theCreative Commons Attribution License (CC BY). Theuse, distribution or reproduction in other forums is per-mitted, provided the original author(s) or licensor arecredited and that the original publication in this journalis cited, in accordance with accepted academic practice.No use, distribution or reproduction is permitted whichdoes not comply with these terms.













doi: 10.3389/fpsyg.2014.00858

Teachers’ expertise in feedback application adapted to thephases of the learning processEva Christophel*, Robert Gaschler and Wolfgang Schnotz

Department of Psychology, Universität Koblenz-Landau, Landau, Germany*Correspondence: [email protected]

Edited by:


Reviewed by:

Craig Speelman, Edith Cowan University, AustraliaMichael H. Connors, Macquarie University, Australia

Keywords: teacher expertise, motivational phases, learning phases, feedback, curriculum scripts

BUILDING BETTER BRAINS—SCHOOLDAY BY SCHOOL DAYIn this essay we highlight feedbackapplication as a domain to study theknowledge and abilities involved in theconstruct of teachers’ expertise. Oneapproach to advance the discussion in afield of research such as expertise acquisi-tion generally is to combine (a) analysis ofcomplex natural phenomena by divisioninto hypothetical simple building blocks,with (b) the synthesis of complex phe-nomena based on known simple buildingblocks (e.g., Braitenberg, 1984; Gaschleret al., 2012). For instance, (a) knowledgerepresentations such as chunks have beenhypothesized to underlie chess expertiseand (b) this analysis in turn was supportedby a computer-run expert system basedon chunks (cf. Lane et al., 2001; Guidaet al., 2012). Gobet et al. (2014) summa-rize research on expert knowledge as wellas the Einstellung effect (e.g., missing a tospot an efficient procedure, because a well-known one is available; cf. Bilalic et al.,2010), and neuro-interventions targetingit. They conclude with the desidera-tum to build better brains—brains thatcan take full advantage of the power ofhypothesis-driven cognition while beingsafeguarded against cognitive illusions andthe Einstellung effect.

In the current article we argue thatbuilding better brains is (an admittedlyto-be-optimized) everyday practice, ratherthan a thought experiment. Like Gobetet al. (2014) we ask how creativity canbe fostered by gaining control over priorknowledge so that it can be flexibly usedor blocked at demand (cf. Bilalic et al.,

2008). Teacher education in universitiesdelivers quasi-experimental conditions forstudying how expertise on learning can beacquired and applied best. In particular,concepts of motivation and action con-trol relevant in robotics and psychologyseem promising in order to capture andstructure the gist of teacher expertise.

BECOMING EXPERTS IN FACILITATINGKNOWLEDGE ACQUISITIONWhile expertise is often studied indomains that yield only hundreds orthousands of experts, universities all overthe world are taking efforts to continu-ously contribute to a large population ofexperts on knowledge acquisition. Basedon theoretical input and years of practice,teachers should be experts for shapingschool lessons such that knowledge acqui-sition is optimized (Bromme, 1997, 2008).Focusing on the demands placed by schoollessons can help to obtain an overviewon the skills teaching experts should have(Bromme, 2008). Teachers have to orga-nize and maintain a structure of studentand teacher activities. This includes theanticipation of students’ inferences as wellas disturbances of the social context oflearning. Teachers need a broad and flex-ible knowledge base covering the subject,as well as a repertoire of instructionalmethods to stimulate students’ learn-ing activities for reaching instructionalobjectives. Like managers, teachers haveto organize the timetable of each lesson tomake sure that the time is mainly used forthe subject matter. Shulman (1986, 1987)attributes to teaching experts a repertoireof content knowledge (e.g., diagnosis of

task specific requirements), pedagogicalcontent knowledge (e.g., how to presentthe subject matter content), and curricu-lar knowledge. Leading to the appropriategeneration and scheduling of feedback tothe students, teaching experts should havea strong background in the philosophy ofthe subject (e.g., subject-related epistemicbeliefs), diagnostic competences (e.g.,judgment of students’ abilities), and skillsthat allow them to juggle between student-related aspects of the learning situation aswell as content-related aspects (Bromme,2008). Teachers’ instructional routinesinclude categorical units called “curricu-lum scripts,” in which subject-relatedand didactic-methodological aspects arelinked. As a consequence of this integratedknowledge, most actions of experiencedteachers proceed automatically (Blömekeet al., 2003). They are difficult to access viaverbal protocols, because in the classroomor a face-to-face situation, verbalizationof the teacher could reduce the learn-ers’ amount of cognitive resources andshift the learner’s attention to the level ofmeta-task processes.

FEEDBACK ALIGNED TOMOTIVATIONAL AND VOLITIONALSTATES OF ACTION REGULATIONLearners face the challenge to acquire andemploy self-regulation strategies in orderto obtain educational outcomes (e.g.,Lerner et al., 2001; Ley and Young, 2001).In order to scaffold learning through adap-tive feedback, teachers need knowledgeabout the dynamics of learning pro-cesses and opportunities to apply feed-back that takes motivational, cognitive













Christophel et al. Teacher expertise

and emotional aspects of learning intoaccount. Apart from finding the appropri-ate frequency for feedback (Healy et al.,2014), we suggest consideration of thematch of feedback type and action phasethe learner is currently in Heckhausen andGollwitzer (1987) have proposed distin-guishing between motivational and voli-tional phases of action regulation. Thedistinction between motivational and voli-tional phases of action regulation has beenused in two different ways to bridge thegap between analysis and synthesis high-lighted by Braitenberg (1984). On theone hand, many architectures of artificialagents in robotics implement the distinc-tion between a motivational phase (theagent is open to new information andready to re-evaluate current preferencesand plans) and a volitional phase (theagent is executing a plan and shieldingitself from novel information that mightlead to a re-evaluation of the plan cur-rently executed; e.g., Visser and Burkhard,2007). On the other hand, cycles of moti-vational and volitional phases are presentin theories of the cyclical and recursivedynamic of learning in educational con-texts (e.g., Zimmerman, 2000; Schmitzand Wiese, 2006).

According to Zimmerman (2000) aswell as Schmitz and Wiese (2006), thelearning process proceeds in three phases.(1) In the goal-setting phase, the learnerchooses goals (e.g., appropriate tasks),plans actions or action steps and selectsadequate strategies. (2) In the perfor-mance phase, the selected strategies haveto be applied in order to complete thetask. (3) In the self-reflection phase, thelearner evaluates his/her learning out-come. Gollwitzer (1990) assumes that ineach one of these phases the learners’attention is focused on specific informa-tion that helps him/her to accomplish thedemands of the task. Taking the currentfocus of the learner into account, feed-back can be better adapted with respect toeffects on the learners’ performance, moodand effort (Ley and Young, 2001; Narciss,2004; Baadte and Schnotz, 2013).

Three types of feedback can be dis-tinguished that align to the respectivephases of the learning process. In thegoal setting phase, the learner shouldreceive goal-setting feedback that informshim/her about how realistic completion of

the chosen task is according to his/her pre-vious performances. In the performancephase, process-feedback offers informa-tion about specific task-inherent demandsand error-related information. In theself-reflection phase, the learner shouldobtain appropriate outcome-feedback thatinforms him/her about the possible causesfor success or failure and about the qual-ity of his/her performance. In contrast,feedback is inappropriate if it does nottake the learner’s phase-specific mind-setinto account and if it does not sup-port the task completion but insteaddecreases the learner’s amount of cognitiveresources available for processing the rel-evant information (e.g., Sweller, 2005; seeChristophel and Baadte, in press).

Matching feedback to the phase-specific requirements of the learningprocess might require flexible strategychanges and adaptation of routines tothe specific content being taught, epis-temic beliefs and motivational/volitionalstate diagnosed. Thus, adaptation andshifting skills seem at least as impor-tant as a large repertoire of strategies.This view is reflected in the notion ofadaptive expertise advocated in educa-tion psychology (Verschaffel et al., 2009;Godau et al., 2014). In contrast to routineexpertise (e.g., expertise that allows theexpert to solve problems very efficientlyand precisely), Hatano and Inagaki (1984)describe adaptive expertise as the potentialto create new solutions and new problemsolving procedures. Adaptive experts are“those who not only perform procedu-ral skills efficiently but also understandthe meaning of the skills and nature oftheir object” (p. 28). In contrast, “routineexperts are outstanding in terms of speed,accuracy, and automaticity of perfor-mance, but lack flexibility and adaptabilityto new problems” (Hatano and Inagaki,1984, p. 31). Verschaffel et al. (2009) havepointed out that flexibility can be definedas the “use of multiple strategies” whileadaptivity includes “making appropriatestrategy choices” (p. 338). This flexibil-ity and adaptivity is necessary in orderto support students’ individualized learn-ing processes with appropriate feedback.Thus, teachers need adaptive expertisewhich should include knowledge about thetask specific requirements, the learner abil-ities, the cyclical and recursive dynamic of

the learning process and the implicationsof this dynamic on motivational, cognitiveand emotional levels. On the one hand,there are first formalized efforts to teachfuture teachers relevant motivational com-petencies (e.g., Rheinberg and Engeser,2010). On the other hand, there are morecautious outlooks as well. According toVerschaffel et al. (2009) “adaptive exper-tise is not something that can be trained ortaught but rather something that has to bepromoted or cultivated” (p. 348).

MORE TEACHING EXPERIENCE DOESNOT NECESSARILY LEAD TO BETTERFEEDBACK SKILLSTeachers’ expertise emerges during thetheoretical and practical phases of teachereducation and professional experienceafter university education (Bromme,2008). It is an open question as to whatextent the professional experience ofteachers promotes and cultivates adaptiveexpertise with regard to adequate stu-dents’ support in individualized lessons.For instance, Christophel (2014) demon-strated that more experienced teachersdid not give more appropriate feedback(i.e., feedback that offers phase-specificinformation) but actually gave more inap-propriate feedback than less experiencedteachers (e.g., feedback that does not sup-port the task completion but reduces thelearner’s amount of cognitive resourcesavailable for processing the relevant infor-mation; Sweller, 2005). Christophel (2014)studied 30 more experienced teachers witha mean professional experience of 11.15years (SD = 12.85) and 30 less experi-enced teachers with a mean professionalexperience of 42.07 days (SD = 89.40).Teachers watched three video-vignettesof students passing through the dif-ferent phases of self-regulated learning(goal-setting, performance and the self-reflection phase). The teachers wereinstructed to stop the films wheneverthey wanted to give feedback to the stu-dents. The results revealed that the moreexperienced teachers stopped the video-vignettes more frequently and more oftenprovided inappropriate feedback to stu-dents as compared to their less experiencedcounterparts. In addition, the study ofBaer et al. (2011) showed that professionalexperience did not lead to better school-ing. Baer and colleagues examined the






development of teacher competences intransition between academic and profes-sional careers. The results demonstratethat teachers with more professional expe-rience did not perform better in themeasured aspects of schooling than grad-uate teachers at the end of their universityeducation.

These findings suggest that professionalexperience alone is not a good predic-tor for the progress of teachers’ exper-tise (Baer et al., 2011; cf. Campitelli andGobet, 2011; Hambrick et al., 2014). Yearsof teaching are possibly not sufficient toinformally and implicitly sensitize mostteachers to cues conveying action phasesand the matching feedback. This couldhave diverse reasons. As explained above,school lessons place numerous demands:teachers have to organize and maintainthe structure of schooling and can beabsorbed by organizational and admin-istrative activities. After university andpractical education, there is a lack ofpeer-coaching (e.g., colleagues who dis-cuss classroom situations) or qualifiedinstruction (e.g., best-practice examples).A lack of (time) resources and motivationto reflect ones feedback behavior mightresult.

ATTENDING THE LARGER PICTUREHowever, training studies can supportan optimistic outlook on the develop-ment of teachers’ expertise in feedbackapplication. Experienced and inexperi-enced teachers can be trained to applyfeedback that supports realistic goal-setting and adequate self-reflection oflearners (Christophel et al., in press).Recommendations helped teachers toincrease feedback in line with the moti-vational phase of the learner from pre-to post-test while phase-inappropriatefeedback could be reduced. Also, atten-tion allocation in the classroom—oftena prerequisite for feedback application—seems to be open to intervention. Forinstance, Miller (2011; cf. Speelman,2014; Wiggins et al., 2014) suggested toemploy eye-tracking in order to provide(becoming) teachers with feedback onhow evenly they distribute their atten-tion across students in the classroom.Students disturbing the setting should notmonopolize attention at the cost of otherstudents. While experienced teachers agree

that uneven distributions of attentionshould be avoided, they lack awareness ofhow well they are yet managing to imple-ment the respective strategies of attentionallocation.

REFERENCESBaadte, C., and Schnotz, W. (2013). Feedback effects

on performance, motivation and mood: are theymoderated by the learner’s self-concept? Scand.J. Edu. Res. 1–22. doi: 10.1080/00313831.2013.781059

Baer, M., Kocher, M., Wyss, C., Guldimann, T.,Larcher, S., and Dörr, G. (2011). Lehrerbildungund Praxiserfahrung im ersten Berufsjahr undihre Wirkung auf die Unterrichtskompetenzenvon Studierenden und jungen Lehrpersonenim Berufseinstieg. [Teachers‘academic careerand their professional experience in the firstyear of their professional career and its impacton their schooling competences]. Zeitschriftfür Erziehungswissenschaft, 14, 85–117. doi:10.1007/s11618-011-0168-5

Bilalic, M., McLeod, P., and Gobet, F. (2008).Inflexibility of experts–reality or myth?Quantifying the Einstellung effect in chessmasters. Cogn. Psychol. 56, 73–102. doi:10.1016/j.cogpsych.2007.02.001

Bilalic, M., McLeod, P., and Gobet, F. (2010). Themechanism of the Einstellung (Set) effect: a per-vasive source of cognitive bias. Curr. Dir. Psychol.Sci. 19, 111–115. doi: 10.1177/0963721410363571

Blömeke, S., Eichler, D., and Müller, C. (2003).Rekonstruktion kognitiver Strukturen vonLehrpersonen als Herausforderung für dieempirische Unterrichtsforschung. Theoretischeund methodologische Überlegungen zuChancen und Grenzen von Videostudien.[Reconstruction of teachers‘cognitive struc-tures as challenge for teaching and learningresearch. Theoretical and methodological consid-erations concerning opportunities and limitationsof video-studies]. Unterrichtswissenschaft,31, 103–121. Available onlineat: http://www.erziehungswissenschaften.hu-berlin.de/institut/abteilungen/didaktik/data/aufsaetze/2003/Methodologie.empirischer.Unterrichtsforschung.pdf

Braitenberg, V. (1984). Vehicles: Experiments inSynthetic Psychology. Cambridge, MA: MIT Press.

Bromme, R. (1997). “Kompetenzen, Funktionenund unterrichtliches Handeln des Lehrers.[Competences, functions and schooling of teach-ers],” in Enzyklopädie der Psychologie. Psychologiedes Unterrichts und der Schule, ed F. E. Weinert(Göttingen: Hogrefe Verlag für Psychologie),177–212.

Bromme, R. (2008). “Lehrerexpertise.[Teachers‘expertise],” in Handbuch derPädagogischen Psychologie, eds W. Schneider andM. Hasselhorn (Göttingen: Hogrefe), 159–167.

Campitelli, G., and Gobet, F. (2011). Deliberate prac-tice: necessary but not sufficient. Curr. Dir. Psychol.Sci. 20, 280–285. doi: 10.1177/0963721411421922

Christophel, E. (2014). Lehrerfeedback imIndividualisierten Unterricht. Spannungsfeld zwis-chen Instruktion und Autonomie. [Teacher Feedbackin Individualized Instruction. Striking the Balance

between Instruction and Autonomy]. Wiesbaden:Springer. doi: 10.1007/978-3-658-05099-3

Christophel, E., and Baadte, C. (in press). “Supportingstudents’ self-regulated learning with teachers’feedback: professional experience as a modera-tor of teachers’ attitude-behavior contingency,”in Multidisciplinary Research on Teaching andLearning, eds W. Schnotz., A. Kauertz, H. Ludwig,A. Müller, and J. Pretsch (Basingstoke, England:Palgrave Macmillan).

Christophel, E., Baadte, C., Heyne, N., and Schnotz,W. (in press). “Diagnostische und didaktischeKompetenz von Lehrkräften zur Förderung derText-/Bild-Integrationsfähigkeit bei Schülerinnenund Schülern der Sekundarstufe I (DIKOL).[Diagnostic and didactic competence of teachersin promoting text-picture-integration abilities ofstudent in lower secondary education (DIKOL)],”in Entwicklung von Professionalität pädagogischenPersonals. Interdisziplinäre Betrachtungen, Befundeund Perspektiven, eds C. Gräsel and K. Trempler(Wiesbaden: Springer).

Gaschler, R., Katz, D., Grund, M., Frensch, P. A.,and Brock, O. (2012). “Intelligent object explo-ration,” in Human Machine Interaction, ed M. Inaki(Rijeka; Shanghai: InTech), 235–260. doi: 10.5772/25836

Gobet, F., Snyder, A., Bossomaier, T., and Harré, M.(2014). Designing a “better” brain: insights fromexperts and savants. Front. Psychol. 5:470. doi:10.3389/fpsyg.2014.00470

Godau, C., Haider, H., Hansen, S., Schubert, T.,and Gaschler, R. (2014). Spontaneously spottingand applying shortcuts in arithmetic – A primaryschool perspective on expertise. Front. Psychol.5:556. doi: 10.3389/fpsyg.2014.00556

Gollwitzer, P. M. (1990). “Action phases and mind-sets,” in Handbook of Motivation and Cognition:foundations of Social Behavior, Vol. 2, eds E. T.Higgins and R. M. Sorrentino (New York, NY:Guilford), 53–92.

Guida, A., Gobet, F., Tardieu, H., and Nicolas, S.(2012). How chunks, retrieval structures andtemplates offer a cognitive explanation for neu-roimaging data on expertise acquisition: a two-stage framework. Brain Cogn. 79, 221–244. doi:10.1016/j.bandc.2012.01.010

Hambrick, D. Z., Oswald, F. L., Altmann, E. M.,Meinz, E. J., Gobet, F., and Campitelli, G.(2014). Deliberate practice: is that all it takesto become an expert? Intelligence 45, 34–45. doi:10.1016/j.intell.2013.04.001

Hatano, G., and Inagaki, K. (1984). Two Courses ofExpertise. Research and Clinical Center for ChildDevelopment, Annual Report, 6. Available online at:http://hdl.handle.net/2115/25206

Healy, A. F., Kole, J. A., and Bourne, L. E. Jr. (2014).Training principles to advance expertise. Front.Psychol. 5:131. doi: 10.3389/fpsyg.2014.00131

Heckhausen, H., and Gollwitzer, P. M. (1987).Thought contents and cognitive function-ing in motivational versus volitional statesof mind. Motiv. Emot. 11, 101–120. doi:10.1007/BF00992338

Lane, P. C. R., Gobet, F., and Cheng, P. C.-H. (2001).What forms the chunks in a subject’s performance?Lessons from the CHREST computational modelof learning. Behav. Brain Sci. 24, 128–129. doi:10.1017/S0140525X01363926


http://www.erziehungswissenschaften.hu-berlin.de/institut/abteilungen/didaktik/data/aufsaetze/2003/Methodologie.empirischer.Unterrichtsforschung.pdf




http://hdl.handle.net/2115/25206




Lerner, R. M., Freund, A. M., De Stefanis, I.,and Habermas, T. (2001). Understanding devel-

opmental regulation in adolescence: the use ofthe selection, optimization, and compensationmodel. Hum. Dev. 44, 29–50. doi: 10.1159/000057039

Ley, K., and Young, D. B. (2001). Instructionalprinciples for self-regulation. Educ. Technol. Res.Dev. 49, 93–103. doi: 10.1007/BF02504930

Miller, K. F. (2011). “Situation awareness in teach-ing - What educators can learn from video-basedresearch in other fields,” in Mathematics TeacherNoticing: Seeing through Teachers’ Eyes, eds M.Sherin, V. Jacobs, and R. Philipp (New York, NY:Routledge), 51–65.

Narciss, S. (2004). The impact of informativetutoring feedback and self-efficacy on motiva-tion and achievement in concept learning. Exp.Psychol. 51, 214–228. doi: 10.1027/1618-3169.51.3.214

Rheinberg, F., and Engeser, S. (2010). “Motive trainingand motivational competence,” in Implicit Motives,eds O. C. Schultheiss and J. C. Brunstein (Oxford,England: Oxford University Press), 510–548.

Schmitz, B., and Wiese, B. (2006). New perspectivesfor the evaluation of training sessions in self-regulated learning: time series-analyses of diarydata. Contemp. Educ. Psychol. 31, 64–96. doi:10.1016/j.cedpsych.2005.02.002

Shulman, L. (1986). Those who understand: knowl-edge growth in teaching. Educ. Res. 15, 4–14. doi:10.3102/0013189X015002004

Shulman, L. (1987). Knowledge and teaching: foun-dations of the new reform. Harv. Educ. Rev.57, 1–22.

Speelman, C. (2014). The acquisition of expertisein the classroom: Are current models of edu-cation appropriate?. Front. Psychol. 5:580. doi:10.3389/fpsyg.2014.00580

Sweller, J. (2005). “Implications of cognitive load the-ory for multimedia learning,” in The CambridgeHandbook of Multimedia Learning, ed R. E. Mayer(New York, NY: Cambridge University Press),19–30.

Verschaffel, L., Luwel, K., Torbeyns, J., and VanDooren, W. (2009). Conceptualizing, investigat-ing, and enhancing adaptive expertise in elemen-tary mathematics education. Eur. J. Psychol. Educ.24, 335–359. doi: 10.1007/BF03174765

Visser, U., and Burkhard, H. D. (2007). RoboCup -10 years of achievements and future challenges. AIMagazine 28, 115–132. doi: 10.1609/aimag.v28i2.2044

Wiggins, M., Brouwers, S., Davies, J., and Loveday,T. (2014). Trait-based cue utilization and initialskill acquisition: implications for models of theprogression to expertise. Front. Psychol. 5:541. doi:10.3389/fpsyg.2014.00541

Zimmerman, B. J. (2000). “Attaining self-regulation.A social cognitive perspective,” in Handbook ofSelf-Regulation, eds M. Boekaerts, P. R. Pintrich,and M. Zeidner (San Diego, CA: Academic Press),13–39.


Received: 30 May 2014; accepted: 19 July 2014;published online: 05 August 2014.Citation: Christophel E, Gaschler R and Schnotz W(2014) Teachers’ expertise in feedback applicationadapted to the phases of the learning process. Front.Psychol. 5:858. doi: 10.3389/fpsyg.2014.00858This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Christophel, Gaschler and Schnotz.This is an open-access article distributed under theterms of the Creative Commons Attribution License(CC BY). The use, distribution or reproduction in otherforums is permitted, provided the original author(s) orlicensor are credited and that the original publicationin this journal is cited, in accordance with acceptedacademic practice. No use, distribution or reproduc-tion is permitted which does not comply with theseterms.













ORIGINAL RESEARCH ARTICLEpublished: 28 November 2014doi: 10.3389/fpsyg.2014.01337

“No level up!”: no effects of video game specialization andexpertise on cognitive performanceFernand Gobet1*, Stephen J. Johnston2, Gabriella Ferrufino3, Matthew Johnston3, Michael B. Jones3,

Antonia Molyneux3, Argyrios Terzis3 and Luke Weeden3

1 Psychological Sciences, University of Liverpool, Liverpool, UK2 Department of Psychology, Swansea University, Swansea, UK3 Department of Psychology, Brunel University, Uxbridge, UK

Edited by:

David Zachary Hambrick, MichiganState University, USA

Reviewed by:

Brooke Noel Macnamara, PrincetonUniversity, USAMark W. Becker, Michigan StateUniversity, USAElizabeth Jane Meinz, SouthernIllinois University Edwardsville, USA

*Correspondence:

Fernand Gobet, PsychologicalSciences, University of Liverpool,Bedford Street South, LiverpoolL69 7ZA, UKe-mail: [email protected]

Previous research into the effects of action video gaming on cognition has suggestedthat long term exposure to this type of game might lead to an enhancement of cognitiveskills that transfer to non-gaming cognitive tasks. However, these results have beencontroversial. The aim of the current study was to test the presence of positive cognitivetransfer from action video games to two cognitive tasks. More specifically, this studyinvestigated the effects that participants’ expertise and genre specialization have oncognitive improvements in one task unrelated to video gaming (a flanker task) andone related task (change detection task with both control and genre-specific images).This study was unique in three ways. Firstly, it analyzed a continuum of expertiselevels, which has yet to be investigated in research into the cognitive benefits ofvideo gaming. Secondly, it explored genre-specific skill developments on these tasksby comparing Action and Strategy video game players (VGPs). Thirdly, it used a verytight experiment design, including the experimenter being blind to expertise level andgenre specialization of the participant. Ninety-two university students aged between 18and 30 (M = 21.25) were recruited through opportunistic sampling and were groupedby video game specialization and expertise level. While the results of the flanker taskwere consistent with previous research (i.e., effect of congruence), there was no effect ofexpertise, and the action gamers failed to outperform the strategy gamers. Additionally,contrary to expectation, there was no interaction between genre specialization and imagetype in the change detection task, again demonstrating no expertise effect. The lackof effects for game specialization and expertise goes against previous research on thepositive effects of action video gaming on other cognitive tasks.

Keywords: change detection task, expertise, flanker task, transfer, video game playing

INTRODUCTIONTransfer—the extent to which skills generalize—is an importanttheoretical concept that has serious practical implications. In aclassic article, Thorndike and Woodworth (1901) propoundedtheir theory of “identical elements,” according to which transferfrom a first domain to a second domain is possible only whenthe components of the skills required in each domain overlap.For example, a pianist can use their knowledge of music theoryto understand a violin concerto, and a mathematician will under-stand the differential equations of an economics paper better thana person without background in mathematics. But even in thesecases, transfer is far from perfect; for example, the pianist will notbe able play the violin concerto itself without extensive additionaltraining.

FAR-TRANSFERIn line with Thorndike and Woodworth’s (1901) hypothesis, mosttheories of expertise predict that transfer from one domain toanother (far-transfer) will be difficult. This is particularly the case

for theories based on the notion that expertise in great part relieson domain-specific perceptual knowledge [e.g., chunking theory(Chase and Simon, 1973) and template theory (Gobet and Simon,1996)]. While perceptual knowledge enables fluid behavior in theoriginal domain, it is of little use in other domains as it doesnot match the new environment. Research on chess has providedconsiderable support for this prediction. Chess players’ percep-tual skills do not extend to visual memory for shapes (Waterset al., 2002), nor do their planning capabilities transfer to theTower of London, a task measuring executive function and plan-ning (Unterrainer et al., 2011). Moreover, contrary to widespreadbelief, there is no robust empirical evidence that playing chessimproves scholastic abilities (Gobet and Campitelli, 2006).

One of the rare domains in which evidence of far transferhas been found is playing action video games (e.g., Green et al.,2009)1. Repeated playing of this kind of game has been reported

1In this study, the term “video game” refers to any published computer or con-sole video game from 1952 to the present day. The term “video game player”













Gobet et al. Player 1 level up!

to lead to improvements in perceptual and attentional processesand to reduce reaction time in other tasks where one must beboth fast and accurate (e.g., Green et al., 2009; Bavelier et al.,2012).

One of the main advantages proposed to be the result ofhabitual action video game playing is that of a more efficientattentional system. For example, Chisholm et al. (2010) com-pared action vs. non-action video game players (VGPs) on anattentional capture task where participants searched for a targetthat could appear in isolation or with a salient task-irrelevantdistractor. Action VGPs showed faster reaction times to detect tar-gets and a reduced effect of distractor interference, leading theauthors to conclude that the action video gamers had better top-down attentional control, with the consequence that they spendless time processing irrelevant distractors. Consistent with thisresult, Hubert-Wallander et al. (2011) found that, compared withnon-action gamers, action gamers demonstrate superior visualselective attention as measured in a visual search task, with thegreatest benefit occurring at the highest cognitive loads (largestsearch arrays). Additional evidence comes from neuroimagingwhere differences in brain activation support the idea that long-term video game playing impacts on cortical functioning. Forexample, using functional magnetic resonance imagery (fMRI),Bavelier et al. (2012) compared a group of action and non-actionvideo gamers on a task involving locating a target stimulus underconditions of increasing distractor load. In addition to overallfaster reaction times, compared to the non-video game playinggroup, the VGPs showed little increase in the level of activationin a network of fronto-parietal sites as distractor load increased.This fronto-parietal network has commonly been associated withattentional processing (Ptak, 2012). These data were taken tosuggest that the VGPs were more efficient in their allocation ofattentional resources such that the cortical sites deploying atten-tion were able to filter out the distracting information moreeasily and therefore showed less load dependent increases incortical activity, supporting the behavioral finding of Chisholmet al. (2010). The proposed superior attentional resource alloca-tion of VGPs (e.g., Hubert-Wallander et al., 2011; Bavelier et al.,2012), that may be at the heart of the observed enhancementof stimulus processing and reduced distractor interference, hasnow been observed in a number of experiments examining videogame expertise based improvements in spatial selective attention(Green and Bavelier, 2003; Feng et al., 2007; Spence et al., 2009),distractor inhibition (Chisholm et al., 2010; Hubert-Wallanderet al., 2011; Mishra et al., 2011; Bavelier et al., 2012), enhancedimage search (Dye et al., 2009) and target detection (Castel et al.,2005; Dye et al., 2009).

As detailed above, a number of reports using different atten-tional tasks have suggested that VGPs have an improved ability to“filter out” unnecessary or irrelevant stimuli partly through theenhancement of attentional functioning (Chisholm et al., 2010;Bavelier et al., 2012). One task that has been used successfully totest VGPs proposed advantage at distractor filtering directly is the

(shortened to VGP) refers to any individual who plays these games, and theterm “non-video game player” (shortened to nVGP) refers to any individualwho partakes in video game play for less than 1 h per week.

Flanker Compatibility Task (Eriksen and Eriksen, 1974; Bavelieret al., 2012). The Flanker Task requires participants to ignoresalient laterally presented distractors while making responses to acentrally presented target stimulus. Mishra et al. (2011) employeda flanker task to examine whether there was any neuroelec-trophysiological evidence of VGP showing enhanced distractorinhibition. The results showed that, behaviourally, VGPs were bet-ter able to ignore flanking items competing for attention witha central stimulus than nVGPs and that this increase in behav-ioral performance was associated with a greater P300 componentin the ERP. The P300 electrophysiological component has beenassociated with perceptual discrimination and decision-making(Picton, 1992; Mishra et al., 2011). These data, together, weretaken to support the hypothesis that VGP were better at filteringout the distractor stimuli leading to improved perceptual deci-sion making. Lavie (1995) also reported that through extensivegame play VGPs gain the ability to identify task-irrelevant flankersbefore further processing stimuli. This indicates that VGPs pos-sess an enhanced capability to logically filter information forrelevance before attempting to ignore distractors, rather thantrying to process everything at once as nVGPs do.

An experimental task that is homologous to the task require-ments of many action video games is the change detection task.In the change detection task, participants are asked to moni-tor a visual display for a small change that they indicate findingvia a keypress response. For example, Clark et al. (2011) foundthat VGPs display a superior ability to spot changes when pre-sented with rapidly alternating sets of images. In this study, 35participants were presented with an unedited/edited image cycleswitching at 4 Hz. The image cycle repeated until the participantsindicated via a mouse click that they had spotted the edited ele-ment by clicking the image in the position they thought believedcontained the image edit. Video game players performed betterthan nVGPs, replicating previous work on attentional improve-ments in VGP (e.g., Green et al., 2009), and there were alsostrategic changes in their search patterns. Compared with nVGP,the VGP showed broader search strategies, further supporting theview that VGPs develop top-down processing.

It follows from the above arguments that, if video-game exper-tise leads to the observed enhanced attentional and perceptualprocessing, then it should be possible to train nVGP using videogames and observe an improvement in their cognitive function-ing. Green and Bavelier (2003) recruited two groups of partici-pants that had no history of video gaming; one group was thentrained on a fast-paced action game (Medal of Honor) and theother on a slow-paced puzzle game (Tetris). After a period of 10 htraining (1 h a day over 10 days), compared with the Tetris group,participants trained on Medal of Honor displayed better UsefulField-of-View (Ball et al., 1988), that is they had an enhance-ment in their ability to search for and identify cued targets. It wasalso found that the Medal of Honor trainees showed a reducedattentional blink (Raymond et al., 1992), i.e., a reduction in thewindow of attentional “blindness” that occurs after detecting orrecognizing the first of two temporally close visual stimuli (Greenand Bavelier, 2003; Feng et al., 2007; Bailey et al., 2010).

While intriguing, the research on the cognitive benefits ofvideo game playing has been criticized on several grounds.






Boot et al. (2011) note that experts and novices have differentexpectations about their performance, which is likely to affecttheir behavior due to demand characteristics. They also observethat playing video games might affect the kind of strategies thatare being used rather than basic perceptual or cognitive capacities.Finally, as some studies have failed to find differences betweenVGPs and nVGPs, the literature might suffer from a file-drawerproblem. Kristjánsson (2013) note that, in many training stud-ies, the control groups do not improve their performance on thetasks of interest, as one would expect, based on the extensive lit-erature on learning, given the test-retest methodology used. Inaddition, the results might be affected by gender differences, as itis difficult to find expert female VGPs. Both Boot et al. (2011) andKristjánsson (2013) note the necessity to carry out independentreplications.

Near transferResearch has also investigated whether transfer occurs betweensub-disciplines of the same field (near transfer). For example,do physicians specializing in neurosurgery generalize their skillswhen solving problems from pediatrics, or do chess players spe-cializing in specific openings (i.e., the first moves of the game)maintain their skill level when making decisions in board posi-tions in which they are not specialized?

Several studies have addressed this issue in medicine (Rikerset al., 2002), political science (Chiesi et al., 1979), and the designof experiments (Schunn and Anderson, 1999). The pattern ofresults suggests that experts fall back on general heuristics whenthey cannot use domain-specific knowledge. Emphasizing therole of general problem-solving methods, these studies also high-light the role of domain-specific patterns and methods, as clearlysome degree of expertise is lost when domain-specific methodsare replaced by domain-general one. While these studies com-pared individuals of the same level of expertise, Bilalic et al. (2009)compared individuals of different skill levels. They took advantageof several features of chess: chess skill is precisely and quantita-tively measured by the Elo rating; chess players enjoy trying tofind the best move in a chess position; and chess players specializein different openings, which makes it relatively easy to find play-ers who have the same strength (as measured by their Elo points)but who have different specialized opening knowledge.

Bilalic et al. compared the performance, in both a memory andproblem solving task, of players who specialized in two differentchess openings. In addition to positions coming from these twotypes of defense, they also used neutral positions (positions dif-ficult to classify with respect to the opening they came from).The players were Candidate Masters, Masters, and InternationalMasters/Grandmasters. The results were dramatic. With only oneexception, all players obtained the best results with the positionstaken from the openings they specialized in. When confrontedwith positions outside their domain of specialization, players per-formed one standard deviation on average below the level shownwith positions taken from their domain of specialization.

AIMS OF THE STUDYMany studies have investigated the differences between VGPsand nVGPs but there is as yet, to our knowledge, no research

establishing whether differing levels of video gaming expertisevary with performance on cognitive tasks. Thus, the first aim ofthis study was to test the hypothesis that, as the level of exper-tise increases, task accuracy increases, and reaction times becomefaster.

In addition, a number of studies compare VGPs who iden-tify as “Action” players to nVGPs, but as yet there has been noresearch into whether the skills demonstrated by action playerscross over into other genres, such as strategy games, or indeedif each genre improves different skills. Data from a consumersurvey by the Entertainment Software Association found thataction and strategy games proved popular with both console andcomputer VGPs, and so these two genres were chosen as a vari-able to test the hypothesis (Entertainment Software Association,2012).

The second aim of the study was thus to test to what extentdifferent video-game genre specialization tap into different cog-nitive abilities. Action games typically involve fast-paced game-play, such as “Call of Duty: Modern Warfare 3” (Infinity Ward,2011), which was the best selling action console video game in2011 (Entertainment Software Association, 2012). It was pre-dicted that the speed of gameplay will heighten action VGPsspeeded response times to stimuli other than those normallyresponded to in a VG, as shown by Green and Bavelier (2003).Strategy players, however, are predicted to possess a strongerreliance on maintaining accuracy as a gained trait from long-term play where accuracy over response time is key to success.This is because, typically, strategy games require the ability tomove and place items in carefully decided places and forma-tions. While often these changes are in response to an in-gametarget event and can result in swift determination and sequenc-ing of new actions to fulfill a shifting long-term goal, there isless emphasis on rapid direct responding to an appearing tar-get. Games such as “StarCraft 2: Wings of Liberty” (BlizzardEntertainment, 2010), the best-selling PC strategy game of 2011(Entertainment Software Association, 2012), demonstrate theneed for this ability, particularly in games with a military basis.As a consequence, it was predicted that strategy players wouldperform with significantly higher mean accuracy and actionplayers would perform with a significantly faster mean reactiontime.

The final aim was to replicate the effect of action video-playingon two tasks: a flanker task (Eriksen and Eriksen, 1974) andchange detection task (Clark et al., 2011). In particular the flankertask has shown a mixed pattern of results; a basic flanker task hasshown both no effect of expertise (Cain et al., 2012), and effects ofexpertise only once an additional perceptual load has been added(Green and Bavelier, 2003). In the case of Green and Bavelier(2003) it was argued that the addition of a perceptual load pre-vented flanker interference in the case of nVGP because, unlikethe VGP, there were fewer spare resources to process the distract-ing flankers. However, it did appear in the original Green andBavelier (2003) that there was a small advantage for VGP com-pared to nVGP at low loads. We therefore predict that, using abasic “low-load” flanker task, there should be a smaller flankereffect for VGP compared to nVGP and that this will increase asVG expertise decreases.





OVERALL METHODETHICAL APPROVALThis study was granted ethical approval from the BrunelUniversity School of Social Sciences ethics board in accordancewith the British Psychological Society (BPS) guidelines. All par-ticipants gave informed consent and were fully debriefed after thestudy.

PILOT STUDYAn online pilot, carried out several months before the main study,asked participants (N = 115) to identify the last three actionand strategy video games they had played. The Call of Duty andAssassin’s Creed video game series were identified as the most pop-ular action video games, and the Starcraft and FIFA series werefound to be the most popular strategy video games.

PARTICIPANTSNinety-two participants (56 male) aged between 18 and 30(M = 21.25, SD = 2.07) were recruited by opportunistic sam-pling through social networking sites and word of mouth. Mostof the participants had filled out the online questionnaire (Pilotstudy). Each participant was offered a food reward for partici-pating in the study, with a further cash reward incentive (£20) ifthey achieved the best score on one of the two tasks out of allparticipants.

APPARATUSThe experiment was run using the E-Prime software package(Psychology Software Tools, Inc., 2008) on a Dell desktop com-puter running Windows 7, with stimuli presented on a 15 inchLenovo LCD screen running at a resolution of 640 × 480 pixels ata refresh rate of 60 Hz. Keyboard and mouse responses were col-lected via a standard keyboard and mouse. Participants were satapproximately 60 cm from the computer screen.

DESIGNThis study was pseudo-experimental in nature, as the indepen-dent variables were not directly manipulated. In both experi-mental measures, the independent variables were skill (experts,intermediates, novices, and controls) and specialization (actionvs. strategy). In some analyses, in order to allow direct compari-son with the literature, we used skill with only two levels (VGPsvs. nVGPs).

In order to operationalise the study variables, criteria for eachbetween-subject variable needed to be established. A question-naire was given at the end of the study. In addition to stan-dard questions such as asking age and gender, information wasobtained on the participants gaming habits to allow for allocationof each participant to the levels of the two independent variables(i.e., VGP or nVGP). The following three questions were asked.

(i) “How many hours a week, on average, do you play videogames for? 0–1, 2–5, 6–10, 11–15, 16–20, 21+.” This ques-tion was used to allocate participants to either the VGP ornVGP group based on their hours of play. Participants whoanswered “0–1” were allocated to the nVGP group and anyanswers above were assigned to the VGP group. Sixty-twoVGPs and 19 nVGPs were identified.

(ii) “On average, what percentage of the games that youplay do you complete? (By completed, we mean attain-ing the highest in-game ‘level’ or ‘rank’ or completingthe game’s storyline. You don’t need to include optionalmissions, achievements or DLC (Downloadable Content).”Participants who answered “76–100%” to this were deemedas Experts (n = 22), those who answered “51–75%” weredeemed Intermediates (n = 24) and those who answered“26–50%” were deemed Novices (n = 22). Those whoanswered “0–25%” were allocated to the Control group(n = 24).

(iii) “Would you identify yourself predominantly as an actionor strategy video gamer?” This question was used to decideeach participant’s genre specialization. Thirty six partici-pants identified themselves as predominantly action VGPs,alongside 32 strategy VGPs.

Table 1 presents the frequency of participants for each genre ×expertise cell of the design.

GENERAL PROCEDUREUpon arrival, one of the research team allocated the participanta random number that corresponded to the participant’s entry inthe pilot database that contained information pertaining to theirgame playing history and experience. They were then handedover to a second experimenter who ran the experiment and whowas blind to the participant’s details and questionnaire scoring.This method ensured that both the participant and the secondresearcher were unaware of the participants’ genre or expertiseallocation. Participants carried out the two experimental mea-sures that formed the study in a random order. For each measure(described below), the first screen to appear was a set of taskinstructions (see below for details of each experiments instruc-tion). Once the series of tasks was complete, participants com-pleted the “General Information Sheet,” wherein they answeredquestions such as age, genre specialization, average weekly hoursplayed etc.

Table 1 | Participation allocation to the experimental groups (genre

specialization and expertise level).

Genre specialization and expertise level Frequency (N)

Action expert 12

Action intermediate 11

Action novice 13

Action control 12

Action total 48

Strategy expert 10

Strategy intermediate 13

Strategy novice 9

Strategy control 12

Strategy total 44

Total 92






DATA ANALYSISOutliers were identified as either reaction times or correctresponses that were notably outside the general distribution.Boxplots for each data set were analyzed and any outliers SPSSidentified were removed. All reaction time analyses were per-formed using correct only trials.

The experiment comprised two measures, an Eriksen Flankertask and a change detection task. Each of these measures isdescribed below.

MEASURE 1: ERIKSEN FLANKER TASKThis measure was a modified version of Eriksen and Eriksen’s(1974) flanker task. Arrows were used instead of letters, similarto other video gaming studies such as Cain et al. (2012).

MATERIALSCongruent and incongruent stimuli were created prior to the startof the experiment by combining arrow stimuli such that the cen-tral arrow to which the participant responded was surroundedby equally spaced, directionally congruent stimuli (e.g., < < <

or > > >) or directionally incongruent stimuli (e.g., < > <

or > < >). The flanker stimuli subtended 8◦ of visual angle.

DESIGNThe experiment was a mixed factorial design with the within sub-jects factor being Congruence and between-subjects factors ofExpertise Level, Genre Affiliation and “Video Game Players vs.Non-Video Game Players.” The dependent variables in this taskwere reaction time and percent correct accuracy.

PROCEDUREParticipants completed 24 congruent and 24 incongruent trials(trial order randomized) in two blocks of equal number (i.e., 24trials per block). On each trial the participant viewed a centrallypresented fixation cross for 500 ms that was replaced by either acongruent or incongruent trial image. Participants viewed eachtrial image and were asked to indicate in which direction thecentral arrow using the arrow keys on the keyboard. Each trialremained onscreen until the participant made a key press. Thenext trial immediately followed.

MEASURE 2: CHANGE DETECTION TASKMATERIALSImages were sourced from Google Image Search and were editedusing Adobe Photoshop CS5 (Adobe Systems, 2010). Based on thepilot study, which provided information on the most commonlyplayed action and strategy games, images from Call of Duty andStarCraft were chosen. As both games are part of a much largerseries of games, the most recent versions of each franchise wereused (StarCraft II: Wings of Liberty; Blizzard Entertainment, 2010)and Call of Duty: Black Ops 2 (Treyarch, 2012). All trial imageswere scaled such that they subtended 26◦ of visual angle.

DESIGNAs it used both expertise and specialization as independent vari-ables, this experiment had the same design as described in Bilalicet al. (2009). Players of different skill levels and specialized withthe video games Call of Duty or StarCraft, as well as a control

group of non-players, were presented with images from thesetwo games in addition to non-video game related (defined as“neutral”) stimuli.

The task itself was an adapted form of Clark et al. (2011).There were 13 trials in total, the first of which was a practicetrial and was not included in later analyses. Three repeated mea-sures conditions were used: Call of Duty, StarCraft and Landscape(Control) with each condition consisting of all those trials con-taining the images derived from those games or scenes. Therewere four images in each condition.

PROCEDUREParticipants fixated a centrally presented cross for 4000 ms priorto the start of the change detection task image presentation. Thefirst, unedited (UE), image was then presented for 240 ms, fol-lowed by a blank gray screen for 80 ms. A second image, identicalto the first, would then appear for a period of 240 ms before beingreplaced by a blank gray screen for 80 ms. The process would thenrepeat but with the edited (E) version (identical save for a changein a single image feature) of the same image, i.e., the sequenceappeared as UE, UE, E, E, UE, UE, E, E. . . This cycle repeateduntil the participant responded by pressing the spacebar on thecomputer keyboard (see Figure 1).

On detection of a change, the participant ceased the trial viakeypress (spacebar) and then reported both the location andnature of the perceived change to the experimenter who remainedwith the participant in the room during data collection. The par-ticipant then pressed the spacebar again in order to trigger thenext trial presentation.

RESULTSMEASURE 1: ERIKSEN FLANKER TASKPrior to analysis, one outlier was removed for failing to complywith task instruction. Data were analyzed using a mixed ANOVA

FIGURE 1 | One cycle of a sample Change Detection trial showing a

Landscape (Control) image. The sequence involved two presentations ofthe unedited image prior to two presentations of an edited image. Anexample of an unedited and associated edited image is shown in thebottom right corner: The feature missing in the edited version of the imageis indicated by the red circle on the original image.





to determine the effect of specialization and skill (between-subjects) with performance on congruent and incongruent trials(within-subjects).

AccuracyAnalyses indicated a main effect of congruence, F(1, 83) = 27.90,p < 0.001, η2

p = 0.25, f = 0.58. The Congruent (Same) condi-tion (M = 23.69, SD = 0.69) had a higher average score thanthe Incongruent (Distractor) condition (M = 18.90, SD = 8.38).This effect was not qualified by participant expertise, F(3, 83) =0.57, p = 0.64, η2

p = 0.02, f = 0.14, or genre, F(1, 83) = 2.55,

p = 0.11, η2p = 0.03, f = 0.18. No interaction was found between

congruence, expertise and genre, F(3, 83) = 2.00, p = 0.12, η2p =

0.07, f = 0.27.

Reaction timeAnalyses showed a main effect of congruence on reaction time,F(1, 73) = 53.31, p < 0.001, η2

p = 0.42, f = 0.85; overall, a lowermean reaction time was demonstrated in the Congruent (Same)condition (M = 433.85, SD = 70.47) than the Incongruent(Distractor) condition (M = 529.46, SD = 155.82). This effectwas not qualified by participant expertise, F(3, 73) = 0.42, p =0.74, η2

p = 0.02, f = 0.14, or genre, F(1, 73) = 1.71, p = 0.59,

η2p = 0.006, f = 0.08. No interaction was found between congru-

ence, expertise and genre, F(3, 73) = 1.71, p = 0.17, η2p = 0.07,

f = 0.27.

VGPs vs. nVGPsIn order to attempt to replicate previous research using this task,we also carried out analyses where the participants were allocatedto only two groups (VGPs and nVGPs). nVGPs were identifiedas any participant who played, on average, less than 1 h of eitherconsole or computer video games per week. Figure 2 shows asummary of the reaction time and accuracy data for the flankertask for VGPs and nVGPs.

With respect to accuracy, there was a main effect of con-gruence on accuracy, F(1, 89) = 19.96, p < 0.001, η2

p = 0.18, f =0.47, that was not qualified by the players’ status [F(1, 89) = 0.24,p = 0.62, η2

p = 0.008, f = 0.09].With respect to reaction time, both groups performed over-

all faster in the Congruent (Same) condition (M = 433.85, SD =70.47) than the Incongruent (Distractor) condition (M = 529.46,SD = 155.82). Analysis indicated a main effect of congruence onreaction time, F(1, 79) = 32.23, p < 0.001, η2

p = 0.29, f = 0.64,that was not qualified by the players’ status, F(1, 79) = 1.12, p =0.29, η2

p = 0.01, f = 0.1.

MEASURE 2: CHANGE DETECTION TASKAfter outliers were removed due to task non-compliance, 85 par-ticipants remained from the original 92. A Mixed ANOVA wascarried out and outliers were controlled in an identical way to theFlanker Task.

Analysis indicated a main effect of image type on reac-tion time, F(2, 154) = 36.57, p < 0.001, n2

p = 0.32, f = 0.69.Response times were quicker in the Landscape condition (M =9399 ms, SD = 5119 ms) than in the Call of Duty condition(M = 16, 138 ms, SD = 7556 ms) and Starcraft condition (M =20, 247 ms, SD = 10, 761 ms). This effect was not qualified byparticipant expertise [F(6, 154) = 1.06, p = 0.39, n2

p = 0.04, f =0.2] or genre [F(2, 154) = 0.57, p = 0.57, n2

p = 0.01, f = 0.1]. Nointeraction was found between image type, expertise and genre[F(6, 154) = 0.27, p = 0.95, n2

p = 0.01, f = 0.1].We also analyzed the data by grouping the participants

into players and non-players. Mauchly’s Test indicated a vio-lated assumption of sphericity, χ2

(2) = 21.57, p < 0.001, thereforedegrees of freedom (df ) were corrected using Greenhouse-Geisserestimates of sphericity (ε = 0.81). Analysis indicated a maineffect of image type on reaction time, F(2, 166) = 25.06, p <

0.001, n2p = 0.23, f = 0.55, that was not qualified by the “Video

Game Players vs. Non-Video Game Players” variable [F(2, 166) =1.37, p = 0.26, n2

p = 0.02, f = 0.14]. Figure 3 shows a summary

FIGURE 2 | Mean accuracy (left), reaction times (right) and the associated 95% confidence intervals for the VGP and nVGP from the Eriksen Flanker

Task.






FIGURE 3 | Mean reaction times along with the associated 95%

confidence intervals for VGP and nVGP from the Change Detection

Task.

of the reaction time data for the change detection task for VGPsand nVGPs.

POWER SUMMARYOne advantage of this study is the relatively large number of par-ticipants who were involved. However, to ensure that the absenceof expertise effects was not due to a lack of statistical power, weinvestigated the size of effects we could be expected to find. Allcalculations are based on a power criterion of 0.8 and a 0.05 alphalevel. For Measure 1: Eriksen Flanker task, one could expect tofind significant differences of effect sizes for the interaction ofcongruency and specialization of 0.31 and of 0.36 for the two-wayinteraction of congruency with skill and for the three-way inter-action of congruency with skill and specialization. For Measure 2:Change Detection Task, effect size of 0.25 could be expected to beidentified for interactions of image type with specialization andeffect sizes of 0.30 for both the two-way interactions of special-ization and skill with image type and the three-way interaction ofimage type, specialization, and skill.

In all cases, the observed effect sizes for game specializationand skill in each of the measures was considerably smaller than theminimum expected detectable effect size even given our relativelylarge sample size. The power analyses also highlight that given oursample size we could expect to detect small effect sizes.

DISCUSSIONOne of the major conclusions of research into learning and exper-tise is that transfer from one domain to another is rare anddifficult, and happens only when the two domains share com-ponents that ask for the same cognitive skills. In recent years, aseries of experiments on action video game playing have found

that playing this kind of game leads to substantial transfer, in par-ticular with tasks engaging attentional processing. The aim of thisstudy was to replicate this phenomenon with two tasks that hadpreviously been used in the action video-game literature. In addi-tion, the study aimed to use a finer measure of expertise than hadbeen done in the past, and to look at the extent to which skillsacquired in a specific VG genre (action or strategy) can be used ina task using material linked to either of the two genres.

In neither task were we able to find any effect of skill or a supe-riority of the VGPs when compared to the nVGPs. Thus, our studydid not support the hypothesis of far transfer, in line with mosttheories of expertise but in contradiction with previous VG stud-ies. Our failure to replicate previous results cannot be ascribed toa lack of power, as the number of participants (n = 92) was highfor this kind of study and our design, incorporating different lev-els of skill, was in principle able to identify subtle skills effects thatcannot be found when only two groups are compared (VGPs vs.nVGPs). In addition, the results we obtained in each task wereconsistent with the results normally obtained in these tasks. Forexample, we found significantly faster reaction times and a greaternumber of percent correct accurate trials for the congruent trialscompared to the non-congruent trials in the flanker task.

For the flanker task, despite strong congruency effects, theabsence of a significantly different interference effect for VGP vs.nVGP is not entirely unexpected, despite our predictions to thecontrary. We argued that based on previous work that we mightexpect a small difference (Green and Bavelier, 2003), especiallygiven our sample size and a finer division of game expertise thanpreviously used. However, this was not the case and, although thenull hypothesis cannot be accepted, the absence of a significantinteraction of game expertise and congruency does add to a num-ber of results showing that, at low loads there is little differencebetween the performance on VGP and nVGP on a flanker task(Green and Bavelier, 2003; Cain et al., 2012).

We followed Boot et al.’s (2011) advice of pre-screening par-ticipants long before the experiments per se, and of asking themto fill in the questionnaire on video game activities at the end ofthe study. Thus, our procedure minimized demand characteris-tics. Together with the fact that other studies have failed to finda VGP effect (e.g., Castel et al., 2005; Boot et al., 2008; Murphyand Spencer, 2009), our results are consistent with the possibilitythat such demands might have played a role in previous researchshowing VGPs’ superiority. However, given that this conclusionis based on null results, further research is needed to validate orinvalidate this hypothesis.

Most unexpected was perhaps the failure to find an expertiseeffect in the change detection task. Near transfer did not occurin spite of the fact that the material used came from the players’domain of expertise (either action or strategy players) and that webased the stimulus choice on the most popular games within eachgenre, an expected level of increased familiarity. Why is it thatthe patterns that the players had presumably acquired by playingtheir favorite game did not enable them to find changes in thestimuli more rapidly? One possible explanation is that the changedetection paradigm is unsuitable for detecting domain-specificpatterns used for unconscious pattern recognition. An examina-tion of the mean reaction times shows that the average time spent





on task was very long (about 20 s on average in the Starcraft con-dition) and thus is likely to engage more conscious mechanisms.This explanation gains further plausibility given that Gobet et al.(in preparation) did find an interaction between specialism andexpertise in a recognition task using a similar design as that usedhere: action players specializing in the Call of Duty and rac-ing players specializing in Gran Turismo performed better whendealing with images from their own game.

Our study was not without weaknesses. The measure of video-game expertise and the allocation to a specific genre were basedon self-reports, and perhaps it would have been desirable (albeitunpractical) to ask players to play segments of their favorite gameto estimate their level. With regard to obtaining a pure mea-sure of specific video game genre benefits a study, such as this,that concentrates on training effects in nVGP (e.g., Green andBavelier, 2003) may be in a better suited to detect subtle changesin performance. A comparison of preferred gametype may beexpected to find subtle differences between players of differentgames but only when those players partake of a single type ofgame. Anecdotally, many game players are “poly-gamers” and willplay other game types in addition to their preferred category. Atraining study where nVGP gain experience playing only a singlegenre of game would therefore be better placed to detect subtledifferences between expertise benefits of individual game types.A similar argument can be made to better examine any potentialgender differences. The sample here did not allow for a meaning-ful investigation of potential gender differences and, for the samereasons previously argued to account for poly-gamers, a trainingstudy would be ideal. Finally, there were few trials in the changedetection task.

In the last year several professional associations and journalshave emphasized the need for more replications. However, in spiteof previous calls (e.g., Gobet et al., 2004), and partly due to thedifficulty of finding experts, research into expertise is rarely repli-cated. The current paper contributes to this effort of obtainingmore robust empirical data.

REFERENCESBailey, K., West, R., and Anderson, C. A. (2010). A negative association between

video game experience and proactive cognitive control. Psychophysiology 47,34–42. doi: 10.1111/j.1469-8986.2009.00925.x

Ball, K., Beard, B., Roenker, D., Miller, R., and Ball, D. (1988). Visual search age andpractice. Invest. Ophthalmol. Vis. Sci. 29, 448.

Bavelier, D., Achtman, R. L., Mani, M., and Föcker, J. (2012). Neural bases ofselective attention in action video game players. Vision Res. 61, 132–143. doi:10.1016/j.visres.2011.08.007

Bilalic, M., McLeod, P., and Gobet, F. (2009). Specialization effectand its influence on memory and problem solving in expert chessplayers. Cogn. Sci. 33, 1117–1143. doi: 10.1111/j.1551-6709.2009.01030.x

Boot, W. R., Blakely, D. P., and Simons, D. J. (2011). Do action video games improveperception and cognition? Front. Psychol. 2:226. doi: 10.3389/fpsyg.2011.00226

Boot, W. R., Kramer, A. F., Simons, D. J., Fabiani, M., and Gratton, G. (2008). Theeffects of video game playing on attention, memory, and executive control. ActaPsychol. 129, 387–398. doi: 10.1016/j.actpsy.2008.09.005

Cain, M. S., Landau, A. N., and Shimamura, A. P. (2012). Action video game experi-ence reduces the cost of switching tasks. Atten. Percept. Psychophys. 74, 641–647.doi: 10.3758/s13414-012-0284-1

Castel, A. D., Pratt, J., and Drummond, E. (2005). The effects of action videogame experience on the time course of inhibition of return and the efficiency

of visual search. Acta Psychol. 119, 217–230. doi: 10.1016/j.actpsy.2005.02.004

Chase, W. G., and Simon, H. A. (1973). Perception in chess. Cogn. Psychol. 4, 55–81.doi: 10.1016/0010-0285(73)90004-2

Chiesi, H. L., Spilich, G. J., and Voss, J. F. (1979). Acquisition of domain-relatedinformation in relation to high and low domain knowledge. J. Verbal Learn.Verbal Behav. 18, 257–273. doi: 10.1016/S0022-5371(79)90146-4

Chisholm, J. D., Kingstone, A., Hickey, C., and Theeuwes, J. (2010). Reduced atten-tional capture in action video game players. Attent. Percept. Psychophys. 72,667–671. doi: 10.3758/APP.72.3.667

Clark, K., Fleck, M. S., and Mitroff, S. R. (2011). Enhanced change detection per-formance reveals improved strategy use in avid action video game players. ActaPsychol. 136, 67–72. doi: 10.1016/j.actpsy.2010.10.003

Dye, M. W. G., Green, C. S., and Bavelier, D. (2009). The development of atten-tion skills in action video game players. Neuropsychologia 47, 1780–1789. doi:10.1016/j.neuropsychologia.2009.02.002

Entertainment Software Association. (2012). Essential Facts Aboutthe Computer and Video Game Industry. Available online at:http://www.theesa.com/facts/pdfs/ESA_EF_2012.pdf (Accessed March 11,2013)

Eriksen, B. A., and Eriksen, C. W. (1974). Effects of noise letters upon identifica-tion of a target letter in a nonsearch task. Percept. Psychophys. 16, 143–149. doi:10.3758/BF03203267

Feng, J., Spence, I., and Pratt, J. (2007). Playing an action video game reducesgender differences in spatial cognition. Psychol. Sci. 10, 850. doi: 10.1111/j.1467-9280.2007.01990.x

Gobet, F., and Campitelli, G. (2006). “Education and chess: a critical review,”in Chess and Education: Selected Essays from the Koltanowski Conference, edT. Redman (Dallas, TX: Chess Program at the University of Texas at Dallas),124–143.

Gobet, F., de Voogt, A. J., and Retschitzki, J. (2004). Moves in Mind. Hove:Psychology Press.

Gobet, F., Patel, K., and Johnston, S., (in preparation). The Effects of Video GameSpecialisation on a Recognition Task.


Green, C. S., and Bavelier, D. (2003). Action video game modifies visual selectiveattention. Nature 423, 534. doi: 10.1038/nature01647

Green, C. S., Li, R. J., and Bavelier, D. (2009). Perceptual learning duringaction video game playing. Topics Cogn. Sci. 2, 202–216. doi: 10.1111/j.1756-8765.2009.01054.x

Hubert-Wallander, B., Sugarman, M., Bavelier, D., and Green, C. S. (2011).Changes in search rate but not in the dynamics of exogenous attention inaction videogame players. Attent. Percept. Psychophys. 73, 2399–2412. doi:10.3758/s13414-011-0194-7

Infinity Ward. (2011). Call of Duty: Modern Warfare 3 [Video Game]. Santa Monica,CA: Activision.

Kristjánsson, Á. (2013). The case for causal influences of action videogameplay upon vision and attention. Attent. Percept. Psychophys. 75, 667–672. doi:10.3758/s13414-013-0427-z

Lavie, N. (1995). Perceptual load as a necessary condition for selective atten-tion. J. Exp. Psychol. Hum. Percept. Perform. 21, 451–468. doi: 10.1037/0096-1523.21.3.451

Mishra, J., Zinni, M., Bavelier, D., and Hillyard, S. A. (2011). Neural basisof superior performance of action videogame players in an attention-demanding task. J. Neurosci. 31, 992–998. doi: 10.1523/JNEUROSCI.4834-10.2011

Murphy, K., and Spencer, A. (2009). Playing video games does not make for bettervisual attention skills. J. Artic. Support Null Hypothesis 6, 1–20.

Picton, T. W. (1992). The P300 wave of the human event-related poten-tial. J. Clin. Neurophysiol. 9, 456–479. doi: 10.1097/00004691-199210000-00002

Ptak, R. (2012). The frontoparietal attention network of the human brain: action,saliency, and a priority map of the environment. Neuroscientist 18, 502–515. doi:10.1177/1073858411409051

Raymond, J. E., Shapiro, K. L., and Arnell, K. M. (1992). Temporary sup-pression of visual processing in an RSVP task—an attentional blink.J. Exp. Psychol. Hum. Percept. Perform. 18, 849–860. doi: 10.1037/0096-1523.18.3.849


http://www.theesa.com/facts/pdfs/ESA_EF_2012.pdf





Rikers, R. M. J. P., Schmidt, H. G., Boshuizen, H. P. A., Linssen, G. C. M., Wesseling,G., and Paas, F. G. W. C. (2002). The robustness of medical expertise: clinicalcase processing by medical experts and subexperts. Am. J. Psychol. 115, 609–629.doi: 10.2307/1423529

Schunn, C. D., and Anderson, J. R. (1999). The generality/specificity of exper-tise in scientific reasoning. Cogn. Sci. 23, 337–370. doi: 10.1207/s15516709cog2303_3

Spence, I., Yu, J. J., Feng, J., and Marshman, J. (2009). Women match men whenlearning a spatial skill. J. Exp. Psychol. Learn. Mem. Cogn. 35, 1097–1103. doi:10.1037/a0015641

Thorndike, E. L., and Woodworth, R. S. (1901). The influence of improvementin one mental function upon the efficiency of other functions. Psychol. Rev. 9,374–382.

Unterrainer, J. M., Kaller, C. P., Leonhart, R., and Rahm, B. (2011). Revisingsuperior planning performance in chess players: the impact of time restrictionand motivation aspects. Am. J. Psychol. 124, 213–225. doi: 10.5406/amer-jpsyc.124.2.0213

Waters, A. J., Gobet, F., and Leyden, G. (2002). Visuo-spatial abilities in chess play-ers. Br. J. Psychol. 30, 303–311. doi: 10.1348/000712602761381402


Received: 31 May 2014; accepted: 03 November 2014; published online: 28 November2014.Citation: Gobet F, Johnston SJ, Ferrufino G, Johnston M, Jones MB, Molyneux A,Terzis A and Weeden L (2014) “No level up!”: no effects of video game specializationand expertise on cognitive performance. Front. Psychol. 5:1337. doi: 10.3389/fpsyg.2014.01337This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Gobet, Johnston, Ferrufino, Johnston, Jones, Molyneux, Terzis andWeeden. This is an open-access article distributed under the terms of the CreativeCommons Attribution License (CC BY). The use, distribution or reproduction in otherforums is permitted, provided the original author(s) or licensor are credited and thatthe original publication in this journal is cited, in accordance with accepted academicpractice. No use, distribution or reproduction is permitted which does not comply withthese terms.













doi: 10.3389/fpsyg.2014.00794

Multi-domain computerized cognitive training programimproves performance of bookkeeping tasks: amatched-sampling active-controlled trialAmit Lampit1,2*, Claus Ebster2,3 and Michael Valenzuela1

1 Regenerative Neuroscience Group, Brain and Mind Research Institute, University of Sydney, Sydney, NSW, Australia2 Lauder Business School, Vienna, Austria3 Department of Marketing, University of Vienna, Vienna, Austria

Edited by:

David Zachary Hambrick, MichiganState University, USA

Reviewed by:

Zach Shipstead, Arizona StateUniversity, USAMichael Francis Bunting, Universityof Maryland Center for AdvancedStudy of Language, USA

*Correspondence:

Amit Lampit, RegenerativeNeuroscience Group, Brain andMind Research Institute, Universityof Sydney, 94 Mallett St.,Camperdown, NSW 2050, Australiae-mail: [email protected]

Cognitive skills are important predictors of job performance, but the extent to whichcomputerized cognitive training (CCT) can improve job performance in healthy adults isunclear. We report, for the first time, that a CCT program aimed at attention, memory,reasoning and visuo-spatial abilities can enhance productivity in healthy younger adultson bookkeeping tasks with high relevance to real-world job performance. 44 businessstudents (77.3% female, mean age 21.4 ± 2.6 years) were assigned to either (a) 20 h ofCCT, or (b) 20 h of computerized arithmetic training (active control) by a matched samplingprocedure. Both interventions were conducted over a period of 6 weeks, 3–4 1-h sessionsper week. Transfer of skills to performance on a 60-min paper-based bookkeeping task wasmeasured at three time points—baseline, after 10 h and after 20 h of training. Repeatedmeasures ANOVA found a significant Group X Time effect on productivity (F = 7.033,df = 1.745; 73.273, p = 0.003) with a significant interaction at both the 10-h (RelativeCohen’s effect size = 0.38, p = 0.014) and 20-h time points (Relative Cohen’s effectsize = 0.40, p = 0.003). No significant effects were found on accuracy or on Conners’Continuous Performance Test, a measure of sustained attention. The results are discussedin reference to previous findings on the relationship between brain plasticity and jobperformance. Generalization of results requires further study.

Keywords: cognitive training, bookkeeping, young adults, job performance, far transfer

INTRODUCTIONCognitive abilities are one of the most significant predictors offuture job performance (Schmidt, 2002). The question thereforearises as to whether interventions that augment cognitive andpsychomotor skills can improve work-related outcomes (Arthuret al., 2003; Academy of Medical Sciences, 2012). Computerizedcognitive training (CCT), which is particularly effective in clini-cal groups and the elderly, is one method for improving cognitiveperformance (Buschert et al., 2010; Vinogradov et al., 2012). Arecent report by leading British scientific societies cited CCT asa potentially effective intervention to improve workplace perfor-mance, stressing the importance of examining the transferabilityof CCT into vocational tasks (Academy of Medical Sciences,2012). Previous studies have reported positive effects of CCT ontasks with high psychomotor demands such as flight (Hart andBattiste, 1992; Gopher et al., 1994), driving (Roenker et al., 2003;Cassavaugh and Kramer, 2009; Pradhan et al., 2011) and laparo-scopic surgery (Schlickum et al., 2009; Adams et al., 2012), aswell as enhanced employment outcomes in schizophrenic patients(Vauth et al., 2005b; McGurk et al., 2007; Bell et al., 2008;Lindenmayer et al., 2008; McGurk et al., 2009). Conversely, arecent RCT conducted by our group (Borness et al., 2013) did notfind any effect of an online CCT program on job performance in

a large cohort of white collar employees of an Australian publicsector organization. Thus, the extent to which CCT can augmentperformance of middle-skill office tasks in young healthy adultsremains unclear. Here, we use a classic example of mid-levelskilled occupational task—bookkeeping—to explore the effec-tiveness of a cognitive training program on work-related taskperformance.

The rapid computerization of accounting practice has trans-formed the nature of contemporary bookkeeping into repetitivetranslation of transactions from natural language into an accoun-tancy software system (Bhaskar et al., 1983; Cooper and Taylor,2000). The objective of basic bookkeeping tasks is to analyzeand organize newly acquired data according to the principles ofdouble entry bookkeeping and subsequently transpose these intojournal entries. The objectivity of these tasks, their prevalencein accounting education, and their relatively low dependencyon external factors (e.g., trends in consumer behavior) makebookkeeping a convenient and occupationally- relevant work-related outcome when studying skill acquisition and the effect ofcognitive factors (Dillard et al., 1982).

An accurate entry of transactions into an accounting system isa bookkeepers’ key performance indicator; therefore, our measureof productivity was based on the number of transactions correctly














Lampit et al. Cognitive training improves bookkeeping performance

transposed. Both speed and accuracy of work are important, asspeed (transactions per unit of time) determines the maximumnumber of transactions that a worker can execute while accuracyrate (the ratio of correct entries out of total transactions) reflectsthe relative frequency of errors. Theoretically, improvements inproductivity could result from an increase in speed and/or accu-racy. However, in practice, increases in task speed usually decreaseaccuracy, a phenomenon known as the speed-accuracy tradeoff(Wickelgren, 1977). When evaluating the effectiveness of any CCTprogram on work-related productivity, the potential influence ofsuch tradeoff must therefore be analyzed (Förster et al., 2003).

Since a CCT intervention typically involves many proceduralfeatures beyond the CCT program itself, which include trainercontact, socialization, motivational prompts, increased stimula-tion, and the traditional Hawthorne expectancy bias and retesteffect, it is vital to compare the putative effects of CCT on pro-ductivity outcomes with respect to an intensity-matched activecontrol (AC) condition to make valid inferences (Jacoby andAhissar, 2013). Yet, of the thirteen trials of CCT in vocational set-tings mentioned above, seven (Gopher et al., 1994; Vauth et al.,2005a; McGurk et al., 2007, 2009; Bell et al., 2008; Cassavaugh andKramer, 2009; Pradhan et al., 2011) did not include an AC arm.We therefore aimed to assess whether CCT can increase the speedor accuracy of work-related task productivity over and above anyeffects seen in an active control condition.

MATERIALS AND METHODSRESEARCH DESIGNThis study was a matched-sampling, double-blind, repeated-measures, active-control trial. Subjects were allocated to either(1) 20 h of CCT arm or (2) active control arm matched in timeand intensity. Performance on a 60-min paper-based bookkeep-ing task (see below) was measured three times, at baseline, after10 h of training, and after 20 h of training. All procedures tookplace in a university classroom converted into a training lab.The study was approved by the Lauder Business School EthicsCommittee.

PARTICIPANTS AND SAMPLING PROCEDUREIn January 2009, all at Lauder Business School (n = 242), anEnglish-language teaching business school in Vienna, Austria,were invited by email to participate in the study. None of the par-ticipants spoke English as first language. Inclusion criteria were(1) successful completion of at least one semester in account-ing, (2) lack of any major neurological or psychiatric disorders,and (3) active participation in a scholarship program offered bythe school. These criteria ensured an adequate understanding ofbasic accounting principles, including the posting task used in thisstudy, and enabled compensation to be offered to volunteers asoutlined below. All candidates gave a written informed consentprior to the baseline assessment.

Of the 44 participants who completed the study, 34 (77.3%)were females and 10 were males. Four subjects (9%) were left-handed. The average age was 21.4 ± 2.6 years. Fourteen partici-pants (31.8%) were 2nd semester students, 15 (34.1%) were 4thsemester students, 14 (31.8%) were 6th semester students, and 1(2.3%) was an 8th semester student. Subjects were compensated

with one “credit” (which provides approximately C40 worth ofdining and housing from the scholarship program) for every hourof participation. The total number of credits was 25, equivalentto about C1000, provided in-kind by the school. Compensationwas offered for participation (in hours) rather than on a pay-per-performance basis. To ensure ethical compensation, participantswho left the study prematurely were compensated based on thenumber of hours performed.

Due to the conservable attentional demand of CCT anddocumented interactions between baseline cognitive ability andresponse to CCT (White and Shah, 2006), we aimed to distributeattentional skill equally between our two groups. The Conners’Continuous Performance Test (Conners, 2000) was adminis-tered to assess sustained attention. Two scores derived from thetest, namely confidence interval and response time variability(Response Time Block Change, RTBC), were used. A lower con-fidence interval indicates better overall attention skills, as doesRTBC as it approaches zero (Homack and Riccio, 2006). Subjectswith the same confidence interval were graded according to theRTBC score and then divided into pairs.

MEASURESPrimary outcome measure: performance on a bookkeeping taskPerformance on a repetitive paper-based posting task was theprimary outcome measure. The task was based on basic rulesof double-entry bookkeeping (Bhaskar et al., 1983) and adaptedfrom didactic tasks included within the students’ curriculum(Kieso et al., 2004). Task materials included: (1) A journal of300 randomly generated transactions, based on 25 different cashand accruals transactions (i.e., payables and receivables) with asum between 1 and 500; (2) A blank general ledger containing14 accounts (assets, liability, equity, income, and expenses), witha summary line every 37 lines of data; and (3) An instructionsheet containing the correct debit and credit postings for each ofthe 25 types of transactions, which was designed to help partic-ipants when not sure how to handle a transaction. Examples ofthe materials used for the measuring bookkeeping performanceare provided in the Supplementary Material.

Three 60-min bookkeeping test sessions were administered, atbaseline (pre-training, T1), after 10 h of training (T2) and after20 h of training (T3). Every test session started with 3 practicetasks to unsure understanding. Participants were asked to workas quickly and as accurately as possible. A performance goal of200 transactions per hour (based on a pilot study) was definedto facilitate self-regulation (Kozlowski and Bell, 2006). A blindedresearch assistant recoded the number of correct debit and creditpostings entries in the corresponding row according to the rulesprovided (productivity) and total entries (to calculate accuracyrate). Only rows in which both postings (debit and credit) wereentered according to the rules were considered as correct.

Secondary measure: sustained attentionIn addition to its role in the matched-sampling process, theConners’ Continuous Performance Test (Conners, 2000) was usedto detect potential intervention effects on participants’ attentionskills. The test was administered at baseline and post-training.The confidence interval, RTBC and Overall (i.e., average) Hit






Reaction Time measures were used (with a lower score indicatingbetter performance).

INTERVENTIONSBoth the experimental (CCT) and active control (AC) groups par-ticipated in 3–4 1-h training sessions weekly for the durationof 6 weeks—the total duration of training in both groups was20 h. Supervised training took place in a computer lab with upto six subjects per session. Supervision aimed to enhance motiva-tion and meta-cognition, according to studies of instruction incognitive training (Gopher et al., 1994; Salas and Burke, 2002;Sandford, 2003; Leemkuil and De Hoog, 2005). The trainingsoftware played sounds through standard speakers, producing adegree of auditory distractions for all participants and increasingselective attentional demands.

Computerized cognitive training (CCT)CCT was based on the commercially available cognitive train-ing software “Captain’s Log” (BrainTrain Inc, Richmond, VA).Captain’s Log has a history of use in clinical populations such asadults diagnosed with traumatic brain injury (Tinius and Tinius,2000; Stathopoulou and Lubar, 2004), schizophrenia (Bellucciet al., 2003), and chronic psychiatric disorders (Burda et al., 1994)as well as children with attention difficulties (Rabiner et al., 2010)and older adults (Eckroth-Bucher and Siberski, 2009).

An adaptive and personalized training program was set usingthe software’s Personal Trainer Wizard, which adjusts training dif-ficulty and content according to predefined settings as well asindividual performance. Visual and auditory distractions wereset at 10% of cases for each type. Seven exercises from theConceptual Memory Skills module (The Ugly Duckling, HappyTrails, Total Recall, Domino Dynamite, Tower Power, Max’sMatch, and What’s Next) and five exercises from the NumericalConcepts/Memory Skills modules (Bits and Pieces, Match Maker,City Lights, Counting Critters, and Happy Hunter) were used.Table 1 describes the cognitive skills trained by each exercise. A

detailed description of each of the exercises is provided in theSupplementary Material.

Active control (AC)Several studies suggest that accounting is associated with strongarithmetic skills, a view that is more likely based on accoun-tancy stereotypes rather than actual demands of the profession(Wells and Fieger, 2006). We therefore chose arithmetic train-ing as our AC condition to control fully for non-specific trainingeffects as well as maintenance of blinding of subjects to the train-ing condition of interest. AC training was based on Maths Trainer,a commercially available computerized arithmetic training pro-gram (Oak Systems, Binstead, UK). The program entails 24 dif-ferent arithmetic exercises (addition, subtraction, multiplication,division, fractions, percentage, number sequence, rounding), aswell as daily tests, kakuro and sudoku. The level of complexityrises as the user progress along the training program. Since a typ-ical training session on Maths Trainer lasts about 45 min, subjectsin the AC group also played Math Ninja (a freeware published byPiotr J. Walczack) a computerized calculation game, for about 10more minutes during each session in order to match the durationof CCT sessions.

ANALYSISAnalyses were perfomed using SPSS 20. Time X Group effectswere tested using repeated measures ANOVA. Where assumptionof sphericity was violated, the Greenhouse-Geisser correction wasapplied to degrees of freedom. Within-group and relative effectsizes were calculated using Cohen’s ds at T2 and T3, using differ-ence from baseline and pooled standard deviation for each timepoint.

RESULTSRECRUITMENT AND PARTICIPANT FLOWParticipant flow is depicted in Figure 1. Overall, 62 studentsexpressed interest to participate. 57 of them were screened

Table 1 | Cognitive skills trained by exercise.

Exercise Skills trained

The Ugly Duckling CPS CR VP WM IM

Happy Trails CR VSQ VPS WM IM SLA

Total Recall CR VSC VPS WM GA IM

Domino Dynamite CR DA VSQ VPS IM

Tower Power CR VSQ VP VT FMC SLA

Max’s Match CPS CR VSC GA IM

What’s Next CPS CR DA VSC VSQ WM FA IM

Bits & Pieces CPS CR VSC VP WM GA IM

Match Maker AA CR VP IM SLA

City Lights CR VSC WM GA IM

Counting Critters CR VSC VSQ VP WM GA IM

Happy Hunter CR VSQ WM FA IM

AA, Alternating Attention; CPS, Central Processing Speed; CR, Conceptual Reasoning; DA, Divided Attention; FA, Focused Attention; FMC, Fine Motor Control; GA,

General Attention; IM, Immediate Memory; SLA, Selective Attention; VP, Visual Perception; VPS, Visual Processing Speed; VS, Visual Scanning; VSC, Visuospatial

Classification; VSQ, Visuospatial Sequencing; VT, Visual Tracking; WM, Working Memory. Detailed description of the exercises can be found in the Supplementary

Material.





for eligibility and were administered the baseline ContinuousPerformance Test; 48 attended an introductory lecture and com-pleted baseline bookkeeping assessment. Subsequently, they wereassigned to groups. Forty four completed the first and secondfollow-up sessions. Of the four participants who left the studyfollowing group assignment (two from each group), three with-drew from the study and one was expelled from the university forreasons unrelated to the study.

BASELINE DATADescriptive statistics and baseline performance are providedin Table 2. Two-sample tests revealed no significant differ-ences between the treatment groups in age, gender, ContinuousPerformance Test scores, productivity, and accuracy. Pearsonzero-order intercorrelations analysis among the study variables atbaseline found a positive correlation between age and attentionalcapacity, as indicated by lower Continuous Performance Testconfidence interval, which implies lower probability of attentiondisorders (r = −0.305, p < 0.05). Males performed worse thandid females on this measure (T = −2.82, p = 0.007). Studentswho had completed more years of their university degree alsoperformed better in baseline bookkeeping (r = 0.446, p < 0.01).There was no relationship between Continuous Performance Testscores and bookkeeping performance.

PRIMARY OUTCOMESTable 3 reports changes in the outcome variables across thethree time points. Repeated measures ANOVA found a significantGroup X Time interaction effect on productivity, as defined by

Table 2 | Baseline measures.

Cognitive training Active control p

(n = 22) (n = 22)

M SD M SD

1. Sexa 72.7 81.8 0.47b

2. Age 21.27 2.51 22.0 2.71 0.36

3. Completed university years 1.27 1.16 1.64 1.18 0.31

4. Productivityc 143.09 50.4 128.91 41.57 0.31

5. Accuracy (%) 91.65 11.71 89.38 7.05 0.44

6. CCPT-II CI 27.07 14.75 28.25 15.69 0.80

7. CCPT-II RTBC -0.021 0.09 -0.02 0.05 0.64

8. CCPT-II Hit RT (msec) 346.46 49.27 352.79 54.07 0.69

aPercentage of females.bPearson X2.cNo. of correct entries in a 60-min session.

CI, Confidence index; RTBS, Response time block change.

FIGURE 1 | Study design and participants flow.






the number of correct entries in a 60-min test session (F = 7.033,df = 1.745; 73.273, p = 0.003), at both the 10-h (Cohen’s effectsize = 0.38, p = 0.014) and 20-h time points (Cohen’s effectsize = 0.40, p = 0.003—see Figure 2). No significant effects werefound for accuracy, although slight improvements of accuracyrates were noted in both the CCT and the AC group (2.61 and4.69%, respectively).

The results indicated strong correlations between performancein these two time points for both productivity (r = 0.65, p <

0.01) and accuracy (r = 0.707, p < 0.01). Whilst the speed of cor-rect entries increased in the CCT group as a result of our training(i.e., increased productivity), we observed no evidence for a corre-sponding drop in accuracy, as accuracy rates at baseline (91.65%)were similar to accuracy at both T2 (95.58%) and T3 (94.32%)(repeated measures F = 0.225, p = 0.638).

SECONDARY OUTCOMESThe results indicated no significant TIME X GROUP interac-tion on either Continuous Performance Test Confidence Index(F = 3.665, p = 0.62), Response Time Block Change (F = 0.663,p = 0.62) or Overall Hit Reaction Time (F = 0.527, p = 0.47, seeTable 3).

DISCUSSIONParticipation in 10 and 20 h of CCT produced significant relativeCohen’s effect sizes (i.e., after accounting for active control train-ing effects) of 0.38 and 0.40 on an untrained bookkeeping task.Because the active control condition included all non-specific fac-tors inherent to training, such as socialization, supervisor inter-action, motivational factors, as well as to some extent cognitivestimulation, the observed effects are unique to CCT and cannotbe explained by test-retest and Hawthorn effects. Moreover, sinceperformance on similar bookkeeping tasks is generally predic-tive of actual bookkeepers’ proficiency (Bhaskar et al., 1983), it isreasonable to hypothesize that this outcome may potentially gen-eralize into real-life job performance in the bookkeeping sector.As CCT may represent a cost-effective type of on-the-job training

for the better promotion of workforce productivity (Academy ofMedical Sciences, 2012), replication of this study in the workplaceenvironment is critical.

We sought to determine whether changes in measures ofattention accompanied and potentially explained our observedbookkeeping effects. None of the Continuous Performance Testmeasures were correlated with either baseline bookkeeping per-formance or changes thereof, and CCT produced no discerniblechange in any of the test measures. One interpretation of thisresult is that sustained attention and inhibitory control has noecological relevance to bookkeeping task performance in healthyyoung adults. Alternatively, a more likely explanation for the lackof effect on response inhibition is that our cognitive training pro-gram did not specifically target this domain, and so the resultingimprovement in bookkeeping skills may be attributed to othercognitive domains. Similarly, we did not observe any difference in

FIGURE 2 | Performance on the 60-min bookkeeping task at the three

time points. Error bars represent standard error. Significant TIME XGROUP differences were observed at T2 and T3.

Table 3 | Estimated marginal means, change rates and summary statistics for outcome variables by training group and assessment time.

Cognitive training Active control Group × Time

(n = 22) (n = 22) Relative effect size

M SD Change, %a M SD Change, %a F p d

Productivity, T2 190 53.88 32.78 150.91 43.91 17.07 6.521 0.014 0.38

Accuracy, T2 (%) 95.58 2.97 4.29 93.24 3.33 4.32 0.001 0.983 −0.24

Productivity, T3 250.73 64.65 75.23 193.5 46.98 50.1 10.226 0.003 0.40

Accuracy, T3 (%) 94.32 4.35 2.91 93.57 3.59 4.69 0.225 0.638 −0.45

CCPT CI, T3 21.77 13.15 −19.61b 26.1 17.21 −7.61b 3.665 0.62 0.25

CCPT RTBC, T3c −0.0005 0.018 −97.62 0.0064 0.016 −1.64 0.663 0.42 0.13

CCPT Hit RT, T3 352.79 26.4 3.37 332.35 56.44 −5.79 0.527 0.47 −0.07

aFrom baseline.bNegative scores mean improvement.cScores closer to zero represent higher attentional performance.

The results in bold highlight significant TIME X GROUP interactions. CCPT, Continuous Performance Test; CI, Confidence index; RTBS, Response time block change;

RT, Response time.





the Continuous Performance Test’s response time measure, whilstparticipants in both groups clearly improved their informationprocessing speed ability on the bookkeeping task.

Therefore, whilst the results indicate that CCT can enhanceperformance on the target bookkeeping task, the nature of theunderlying cognitive mediators remain unclear. Lack of addi-tional cognitive measures to probe for these mechanisms is there-fore a major limitation of this study, and future studies shouldcomplement functional outcomes with a wider neuropsycholog-ical battery. What cognitive changes could theoretically medi-ate the observed improvement in bookkeeping performance?Previous task analyses in the field of bookkeeping (Dillard et al.,1982; Bhaskar et al., 1983; Dillard, 1984) point to several cog-nitive processes that may be taxed by bookkeeping tasks andthus underpin skill acquisition. Typically, bookkeeping involvesintegrating new information (a particular transaction) with pre-vious knowledge of how to categorize this information withinthe bookkeeping set of accounts. Working memory and rule-based decision-making may therefore play an important role inbookkeepers’ ability to quickly and accurately classify and recordtransactions (Dillard, 1984). It is also clear that participants inthe CCT group spent less time on each individual transactioncompared to controls, suggesting a CCT-induced improvementin information processing speed. Indeed, of the 12 exercises inour CCT regimen, 8 (66%) had a working memory component,7 (58%) had a processing speed component, and all had a con-ceptual reasoning component at increasing rate of difficulty (seeTable 1 and Supplementary Material for descriptions of the exer-cises and the cognitive skills trained by each). However, transferto these domains was not formally tested and remains an openarea for further research.

Our study has several other limitations. First, the sample sizewas quite small and warrants replication. Second, the results didnot identify cognitive skills that are essential to bookkeeping per-formance or factors (e.g., age, experience, and compensation)that may predict response to training. Third, the subjects were ahighly-educated, young, and relatively-restricted university-basedvolunteer cohort, unlike the heterogeneous workforce. Subjectswere also highly motivated to complete the training (and did so)for secondary gains. Whether workers in a busy work environ-ment facing multiple time- and productivity- pressures wouldrespond similarly is debatable, and the dollar value of CCT toorganizations is unknown.

To that end, we have recently reported negative resultsfrom a larger trial of CCT in organizational settings (Bornesset al., 2013). This trial was based on short (15–20 min) self-administered sessions, whereas the current study provided 60-min group-based, supervised sessions. Supervision and supportmay therefore be crucial for training success (Salas and Burke,2002; Sandford, 2003; Medalia and Richardson, 2005), and wasassociated with greater effects in healthy older adults (Kelly et al.,2014). Other differences between these two studies, most notablythe use of distinct CCT programs and choice of productivity-related endpoints emphasize the need for more specific protocolsin this area.

Improving workforce cognition along the life span may be akey factor in maintaining economic prosperity in the context of

ageing populations and growing weight of cognitively demand-ing tasks in the contemporary workplace. Though limited inscope, our laboratory-based study suggests that supervised CCTcan help improve performance on a work-related task such asbookkeeping. Field-based research in workplace settings is nowrequired to test the idea that CCT can boost workers’ cogni-tive abilities and translate to enhanced real-world occupationaloutcomes such as work productivity.

ACKNOWLEDGMENTSAmit Lampit is supported by the Dreikurs Bequest. MichaelValenzuela is a National Health and Medical Research Councilof Australia research fellow (ID 1004156), and received fund-ing from the Brain Department LLC as well as in-kind supportfrom BrainTrain Inc for projects unrelated to this study. Wethank the study participants as well as Silvia Kucera, MartinSamek, Alla Fayvishenko, and Eva Hammerschmid for their helpin conducting the study.


00794/abstract

REFERENCESAcademy of Medical Sciences. (2012). Human Enhancement and the Future

of Work [Online]. Available online at: http://royalsociety.org/uploadedFiles/Royal_Society_Content/policy/projects/human-enhancement/2012-11-06-Human-enhancement.pdf [Accessed 2 July 2014].

Adams, B. J., Margaron, F., and Kaplan, B. J. (2012). Comparing video games andlaparoscopic simulators in the development of laparoscopic skills in surgicalresidents. J. Surg. Educ. 69, 714–717. doi: 10.1016/j.jsurg.2012.06.006

Arthur, W. Jr., Bennett, W. Jr., Edens, P. S., and Bell, S. T. (2003). Effectivenessof training in organizations: a meta-analysis of design and evaluation features.J. Appl. Psychol. 88, 234–245. doi: 10.1037/0021-9010.88.2.234

Bell, M. D., Zito, W., Greig, T., and Wexler, B. E. (2008). Neurocognitive enhance-ment therapy with vocational services: work outcomes at two-year follow-up.Schizophr. Res. 105, 18–29. doi: 10.1016/j.schres.2008.06.026

Bellucci, D. M., Glaberman, K., and Haslam, N. (2003). Computer-assisted cog-nitive rehabilitation reduces negative symptoms in the severely mentally ill.Schizophr. Res. 59, 225–232. doi: 10.1016/S0920-9964(01)00402-9

Bhaskar, R., Dillard, J. F., and Stephens, R. G. (1983). Skill acquisitionin the bookkeeping task. Instr. Sci. 11, 329–356. doi: 10.1007/BF00137293

Borness, C., Proudfoot, J., Crawford, J., and Valenzuela, M. (2013). Puttingbrain training to the test in the workplace: a randomized, blinded, multi-site, active-controlled trial. PLoS ONE 8:e59982. doi: 10.1371/journal.pone.0059982

Burda, P. C., Starkey, T. W., Dominguez, F., and Vera, V. (1994). Computer-assistedcognitive rehabilitation of chronic psychiatric inpatients. Comput. Hum. Behav.10, 359–368. doi: 10.1016/0747-5632(94)90061-2

Buschert, V., Bokde, A. L. W., and Hampel, H. (2010). Cognitive intervention inAlzheimer disease. Nat. Rev. Neurol. 6, 508–517. doi: 10.1038/nrneurol.2010.113

Cassavaugh, N. D., and Kramer, A. F. (2009). Transfer of computer-based train-ing to simulated driving in older adults. Appl. Ergon. 40, 943–952. doi:10.1016/j.apergo.2009.02.001

Conners, C. K. (2000). Conners’ Continuous Performance Test II. Toronto, ON:Multi-Health Systems.

Cooper, C., and Taylor, P. (2000). From Taylorism to Ms Taylor: the transformationof the accounting craft. Account. Org. Soc. 25, 555–578. doi: 10.1016/S0361-3682(99)00052-5

Dillard, J. F. (1984). Cognitive science and decision making research in accounting.Account. Org. Soc. 9, 343–354. doi: 10.1016/0361-3682(84)90018-7




http://royalsociety.org/uploadedFiles/Royal_Society_Content/policy/projects/human-enhancement/2012-11-06-Human-enhancement.pdf







Dillard, J. F., Bhaskar, R., and Stephens, R. G. (1982). Using first-order cognitiveanalysis to understand problem solving behavior: an example from accounting.Instr. Sci. 11, 71–92. doi: 10.1007/BF00120982

Eckroth-Bucher, M., and Siberski, J. (2009). Preserving cognition through an inte-grated cognitive stimulation and training program. Am. J. Alzheimer’s Dis. OtherDemen. 24, 234–245. doi: 10.1177/1533317509332624

Förster, J., Higgins, E. T., and Bianco, A. T. (2003). Speed/accuracy decisions in taskperformance: built-in trade-off or separate strategic concerns? Organ. Behav.Hum. Decis. Process. 90, 148–164. doi: 10.1016/S0749-5978(02)00509-5

Gopher, D., Weil, M., and Bareket, T. (1994). Transfer of skill from a computergame trainer to flight. Hum. Factors 36, 387–405.

Hart, S. G., and Battiste, V. (1992). Field test of video game trainer. Hum. FactorsErgon. Soc. Annu. Meet. Proc. 36, 1291–1295. doi: 10.1518/107118192786749450

Homack, S., and Riccio, C. A. (2006). Conners’ Continuous Performance Test(2nd ed.; CCPT-II). J. Atten. Disord. 9, 556–558. doi: 10.1177/1087054705283578

Jacoby, N., and Ahissar, M. (2013). What does it take to show that a cognitive train-ing procedure is useful? A critical evaluation. Prog. Brain Res. 207, 121–140. doi:10.1016/B978-0-444-63327-9.00004-7

Kelly, M. E., Loughrey, D., Lawlor, B. A., Robertson, I. H., Walsh, C., and Brennan,S. (2014). The impact of cognitive training and mental stimulation on cognitiveand everyday functioning of healthy older adults: a systematic review and meta-analysis. Ageing Res. Rev. 15C, 28–43. doi: 10.1016/j.arr.2014.02.004

Kieso, D. E., Weygandt, J. J., and Warfield, T. D. (2004). Intermediate Accounting.Hoboken, NJ: John Wiley & Sons.

Kozlowski, S. W. J., and Bell, B. S. (2006). Disentangling achievement orienta-tion and goal setting: effects on self-regulatory processes. J. Appl. Psychol. 91,900–916. doi: 10.1037/0021-9010.91.4.900

Leemkuil, H., and De Hoog, R. (2005). “Is support really necessary within edu-cational games?” in Workshop on Educational Games as Intelligent LearningEnvironments, 12th International Conference on Artificial Intelligence inEducation (AIDE 05), eds C. Conati and S. Ramachandran (Amsterdam), 21–31.

Lindenmayer, J. P., McGurk, S. R., Mueser, K. T., Khan, A., Wance, D., Hoffman,L., et al. (2008). A randomized controlled trial of cognitive remediation amonginpatients with persistent mental illness. Psychiatr. Serv. 59, 241–247. doi:10.1176/appi.ps.59.3.241

McGurk, S. R., Mueser, K. T., Derosa, T. J., and Wolfe, R. (2009). Work, recovery,and comorbidity in schizophrenia: a randomized controlled trial of cognitiveremediation. Schizophr. Bull. 35, 319–335. doi: 10.1093/schbul/sbn182

McGurk, S. R., Mueser, K. T., Feldman, K., Wolfe, R., and Pascaris, A.(2007). Cognitive training for supported employment: 2–3 year outcomesof a randomized controlled trial. Am. J. Psychiatry 164, 437–441. doi:10.1176/appi.ajp.164.3.437

Medalia, A., and Richardson, R. (2005). What predicts a good responseto cognitive remediation interventions? Schizophr. Bull. 31, 942–953. doi:10.1093/schbul/sbi045

Pradhan, A., Divekar, G., Masserang, K., Romoser, M., Zafian, T., Blomberg,R. D., et al. (2011). The effects of focused attention training on the dura-tion of novice drivers’ glances inside the vehicle. Ergonomics 54, 917–931. doi:10.1080/00140139.2011.607245

Rabiner, D., Murray, D., Skinner, A., and Malone, P. (2010). A randomized trial oftwo promising computer-based interventions for students with attention diffi-culties. J. Abnorm. Child Psychol. 38, 131–142. doi: 10.1007/s10802-009-9353-x

Roenker, D. L., Cissell, G. M., Ball, K. K., Wadley, V. G., and Edwards, J. D. (2003).Speed-of-processing and driving simulator training result in improved drivingperformance. Hum. Factors 45, 218–233. doi: 10.1518/hfes.45.2.218.27241

Salas, E., and Burke, C. S. (2002). Simulation for training is effective when. Qual.Saf. Health Care 11, 119–120. doi: 10.1136/qhc.11.2.119

Sandford, J. A. (2003). “Cognitive training and computers: an innovativeapproach,” in Therapist’s Guide to Learning and Attention Disorders, eds H. F.Aubrey and A. K. Ronald (San Diego, CA: Academic Press), 421–441.

Schlickum, M. K., Hedman, L., Enochsson, L., Kjellin, A., and Fellander-Tsai, L.(2009). Systematic video game training in surgical novices improves perfor-mance in virtual reality endoscopic surgical simulators: a prospective random-ized study. World J. Surg. 33, 2360–2367. doi: 10.1007/s00268-009-0151-y

Schmidt, F. L. (2002). The role of general cognitive ability and job perfor-mance: why there cannot be a debate. Hum. Perform. 15, 187–210. doi:10.1080/08959285.2002.9668091

Stathopoulou, S., and Lubar, J. F. (2004). EEG changes in traumatic braininjured patients after cognitive rehabilitation. J. Neurother. 8, 21–51. doi:10.1300/J184v08n02_03

Tinius, T. P., and Tinius, K. A. (2000). Changes after EEG biofeedback and cogni-tive retraining in adults with mild traumatic brain injury and attention deficithyperactivity disorder. J. Neurother. 4, 27–44. doi: 10.1300/J184v04n02_05

Vauth, R., Corrigan, P. W., Clauss, M., Dietl, M., Dreher-Rudolph, M., Stieglitz,R. D., et al. (2005a). Cognitive strategies versus self-management skillsas adjunct to vocational rehabilitation. Schizophr. Bull. 31, 55–66. doi:10.1093/schbul/sbi013

Vauth, R., Corrigan, P. W., Clauss, M., Dietl, M., Dreher-Rudolph, M., Stieglitz,R. D., et al. (2005b). Cognitive strategies versus self-management skillsas adjunct to vocational rehabilitation. Schizophr. Bull. 31, 55–66. doi:10.1093/schbul/sbi013

Vinogradov, S., Fisher, M., and De Villers-Sidani, E. (2012). Cognitive training forimpaired neural systems in neuropsychiatric illness. Neuropsychopharmacology37, 43–76. doi: 10.1038/npp.2011.251

Wells, P. K., and Fieger, P. (2006). High school teachers’ perceptions of accounting:an international study. Aust. J. Account. Educ. 2, 29–51.

White, H. A., and Shah, P. (2006). Training attention-switching ability in adultswith ADHD. J. Atten. Disord. 10, 44–53. doi: 10.1177/1087054705286063

Wickelgren, W. A. (1977). Speed-accuracy tradeoff and information processingdynamics. Acta Psychol. 41, 67–85. doi: 10.1016/0001-6918(77)90012-9

Conflict of Interest Statement: Dr. Valenzuela received funding from the BrainDepartment LLC as well as in-kind support from BrainTrain Inc for projectsunrelated to this study.Received: 17 February 2014; accepted: 06 July 2014; published online: 28 July 2014.Citation: Lampit A, Ebster C and Valenzuela M (2014) Multi-domain computerizedcognitive training program improves performance of bookkeeping tasks: a matched-sampling active-controlled trial. Front. Psychol. 5:794. doi: 10.3389/fpsyg.2014.00794This article was submitted to Cognition, a section of the journal Frontiers inPsychology.Copyright © 2014 Lampit, Ebster and Valenzuela. This is an open-access article dis-tributed under the terms of the Creative Commons Attribution License (CC BY). Theuse, distribution or reproduction in other forums is permitted, provided the originalauthor(s) or licensor are credited and that the original publication in this jour-nal is cited, in accordance with accepted academic practice. No use, distribution orreproduction is permitted which does not comply with these terms.













doi: 10.3389/fpsyg.2014.00762

On the effect of chess training on scholastic achievementWilliam M. Bart*

Department of Educational Psychology, University of Minnesota, MN, USA*Correspondence: [email protected]

Edited by:


Reviewed by:

Philippe Chassy, Liverpool Hope University, UK

Keywords: chess, cognition, education, scholastic achievement, expertise

What are the effects of chess training—especially on scholastic achievementamong school-aged students? Can chessinstruction facilitate the acquisition ofscholastic competency? The current stateof the research literature is that chesstraining tends not to provide educationalbenefits. This article provides a criticalreview of research on the effects of chesstraining on the scholastic achievementlevels of school-aged students.

EDUCATIONAL BENEFITS OF CHESSVarious studies and conference presenta-tions (e.g., Christiaen and Verholfstadt,1978; Liptrap, 1998; Bart and Atherton,2004) provided results in support of theeducational benefits of chess instruction inthe schools. Gobet and Campitelli (2006)reviewed that research and reached thefollowing conclusions: (a) the educationaleffects of optional scholastic chess train-ing remain undetermined; (b) compulsoryscholastic chess instruction may engendermotivational problems among students;and (c) chess instruction may be bene-ficial among novices, but is less impor-tant among intermediate and advancedplayers for whom the amount of prac-tice and the acquisition of knowledge areof paramount importance. Gobet et al.(2004) contended that such conclusionsare in line with the view of de Groot(1977, 1978) that educational benefitsof chess instruction are likely “low-levelgains” such as improvements in attentionand concentration and interest in learn-ing, rather than “high-level gains” suchas improvements in intelligence, scholasticachievement, and creativity.

Additional research is supportive of deGroot’s view. Bilalic et al. (2007) deter-mined that intelligence explained a smaller

amount of the variance in chess skillamong competent young chess playersthan the amount of practice time. Waterset al. (2002) also found little support fora relationship between intelligence andchess skill.

Contrary to the results of Gobet andCampitelli (2006) that chess instructionprovides very modest if any educationalbenefits is research that attests to the bene-fits of chess training. For example, Smithand Cage (2000) reported the effects of120 h of chess instruction on the mathe-matics achievement among rural, African-American secondary school students innorthern Louisiana. They determined thatthe treatment group composed of 11females and 10 males scored significantlyhigher in mathematics achievement andnon-verbal cognitive ability than the con-trol group composed of 10 females and10 males after controlling for differencesamong pretest scores.

In a more recent study Aciego et al.

(2012) used a quasi-experimental studyto examine the cognitive effects of chesstraining. The experimental group con-sisted of 170 students, 6–16 years ofage, who received extracurricular chessinstruction. The comparison group con-sisted of 40 students in a similar age range.Those students received extracurricularsports (soccer or basketball) activities. TheWechsler Intelligence Scale for Children(WISC-R) and a record completed by thetutor-teacher to measure problem solvingwere dependent variables.

After adjusting for pretest scores, thechess group registered significantly higherposttest scores than the sports group forfive of nine WISC-R subtests—i.e., theSimilarities, Digit Span, Block Design,Object Assembly, and Mazes subtests. The

chess group also registered significantlyhigher posttest scores in problem solvingcapacity than the sports group. Theauthors concluded that chess is a “valuableeducational tool” (p. 558).

In another recent study Kazemi et al.(2012) examined the cognitive effectsof chess play. They employed an experi-mental group composed of 86 randomlyselected school-aged students, whoreceived chess instruction for six months,and a control group of 94 randomlyselected school-aged students. All par-ticipants were male and from 5th, 8th, and9th grades from schools in Shanandaj inwestern Iran. All participants were admin-istered a measure of metacognitive abilityand a grade-appropriate mathematicsexam prior to and after the intervention.

The chess group participants registered

significantly higher posttest metacognitiveability scores and higher posttest math-ematics test scores than the non-chessgroup participants. A major conclusionof the study is that chess instructionimproves significantly the mathematicalabilities and the metacognitive capacitiesof school-aged students.

In a third study, Trinchero (2013)examined the effects of chess instructionon the mathematical ability of primaryschool students. His study involved 568primary school children in Italy placed infour groups: (1) experimental, (2) control,(3) experimental without a pretest, and (4)control without a pretest. The experimen-tal group received chess training in addi-tion to ordinary class lessons. The controlgroup only received ordinary class lessons.One prominent result was that the exper-imental group that received chess trainingregistered a modest but statistically signifi-cant increase in scores on mathematics test












Bart On the effect of chess training on scholastic achievement

items that required problem-solving skillson complex tasks. That effect was greateramong students who had more hours ofchess instruction.

These last four studies lend supportto the view that chess training has pos-itive cognitive effects on regular school-aged students. In addition, there are somestudies that address the issue of cognitiveeffects of chess training on school-agedstudents with disabilities.

Scholz et al. (2008) investigated theeffects of chess training on mathemat-ics learning among students with learningdisabilities based on intelligence scores inthe 70–85 IQ range. School classes fromfour elementary schools in Germany wererandomly assigned to two groups: (a) anexperimental group that received chessinstruction of one hour per week for oneentire school year; and (b) a comparisongroup that received supplementary math-ematics instruction of one hour per week.The two groups did significantly differ intheir calculation abilities for simple addi-tion tasks and counting. The authors con-cluded “chess could be a valuable learningaid for children with learning disabilities”(p. 138).

In a second study Barrett and Fish(2011) investigated the cognitive effects ofa 30-week chess-training program withinmathematics classes for students in specialeducation in a middle school in southwest-ern United States. All participants qual-ified for special education services andwere in either 6th, 7th, or 8th grades.A sample of 31 participants were ran-domly placed into two groups: (a) anexperimental group composed of 15 stu-dents who received the chess instruc-tion along with a sizable portion of theregular instruction in resource mathe-matics specially designed for students inspecial education; and (b) a compari-son group composed of 16 students whoreceived all of the regular mathematicsinstruction in resource mathematics ratherthan any chess instruction. The depen-dent variables for this study were scoreson the mathematics Texas Assessment ofKnowledge and Skills (TAKS) that is astandardized test of mathematics compe-tencies for pre-collegiate students in Texas,and end-of-year course grades in resourcemathematics. All participants completeda version of the mathematics TAKS

that was modified for special educationstudents.

This study had some interesting results.First, there was a significant relationshipbetween chess instruction and end-of-year grades. Second, there was a statisti-cally significant relationship between chessinstruction and mathematics TAKS testscores.

This study provided support that chessinstruction facilitates transfer of cogni-tive skills from chess to mathematics forstudents in special education.

These latter two studies using spe-cial education students indicate that chessinstruction has the potential to promotemathematics achievement among studentsin special education.

In a third study Hong and Bart (2007)examined the cognitive effects of chessinstruction on students at risk for aca-demic failure in Korea. The total sampleof participants was 38 students from threeelementary schools randomly placed intotwo groups: (a) an experimental groupthat received 90-min chess lessons weeklyfor 3 months; and (b) a comparison groupthat received regular school activities afterclass.

All participants were tested prior toand after the intervention with the fol-lowing several instruments: (a) the Testof Nonverbal Intelligence—Third Edition(TONI-3) to measure cognitive ability ina language-free manner; and (b) a ChessSkill Rating method using chess softwareto assess level of chess competency.

TONI-3 posttest scores and chess skillratings were significantly correlated aftercontrolling for TONI-3 pretest scores.This statistical finding suggests that chessinstruction that produces higher chess skillratings may lead to gains in levels of non-verbal intelligence among students at riskfor academic failure.

These studies indicating positive effectsof chess training among students with dis-abilities support the view of Storey (2000)who extolled the use of chess training asa means to promote higher-order think-ing skills among disabled students. Storey(2000, p. 47) recommended that teachersconsider “chess as an instructional strat-egy for reinforcing skills such as concen-tration, problem identification, problemsolving, planning strategies, creativity, andlucid thinking.”

But why would chess training lead toimprovements in scholastic achievement?To play chess well, one must attend to andcomprehend chess positions and inducepatterns among the pieces, an indication offluid intelligence and concentration capac-ity. The chess positions can be very com-plex with up to 32 pieces from six piecetypes arrayed on a 64 square board.

One must then formulate and evalu-ate possible moves, an indication of exec-utive functioning and critical thinking.For example, middle game positions oftenpermit 30 different legal moves at everyturn. The chess player must ideally eval-uate positions resulting from such movesselecting the move that produces the posi-tion most advantageous to the player. Thechess player must evaluate chess movesand their resulting positions without actu-ally moving any pieces. There are thussubstantial demands on visual workingmemory.

In chess, one must engage in thissequence of (1) position comprehension,(2) pattern induction, and (3) move for-mulation and evaluation relatively quickly.This coordinated set of cognitive skillsrequired in competent chess play likelytransfers to the learning of mathematicsand related fields that also often requirecomprehension, induction, analysis, andevaluation of complex phenomena. Thisconstitutes a theoretical framework whychess training likely has cognitive benefits.This cognitive explanation for the benefitsof chess training is compatible with com-parable explanations provided by Storey(2000) and Trinchero (2013).

CHESS TRAINING, EXPERTISE, ANDTHE ISSUE OF RESEARCH RIGORThe research reported thus far providesevidence that chess training has salutarycognitive and educational effects amongschool-aged students. However, the argu-ment of Gobet and Campitelli (2006)needs to be considered before we canbe confident that chess training is avalid means to improve scholastic achieve-ment levels. To Gobet and Campitelli(2006), rigorous experimental research isneeded to determine the extent to whichchess training has strong cognitive andeducational effects.

However, such rigorous experimen-tal inquiry involving, for example,





Bart On the effect of chess training on scholastic achievement

random placement of participants intoexperimental and control groups is costlyand difficult to implement in a school set-ting. What is needed is an increase in thequality and quantity of empirical stud-ies to determine the extent to which theacquisition of chess expertise facilitates theacquisition of scholastic expertise amongstudents.

REFERENCESAciego, R., Garcia, L., and Betancourt, M. (2012).

The benefits of chess for the intellectual andsocial-emotional enrichment in schoolchil-dren. Span. J. Psychol. 15, 551–559. doi:10.5209/rev_SJOP.2012.v15.n2.38866

Barrett, D., and Fish, W. (2011). Our move: usingchess to improve math achievement for studentswho receive special education services. Int. J. Spec.Educ. 26, 181–193.

Bart, W., and Atherton, M. (2004). “The neurosci-entific basisof chess playing: applications to thedevelopment of talent and education,”in PaperPresented at the “Learning to Know the BrainConference” (Amsterdam).

Bilalic, M., McLeod, P., and Gobet, F. (2007).Does chess need intelligence? – A study withyoung chess players. Intelligence 35, 457–470. doi:10.1016/j.intell.2006.09.005

Christiaen, J., and Verholfstadt, D. C. (1978). Chessand cognitive development. Nederlandse Tydschriftvoorde Psychologie en haar Grensegebieten, 36,561–582.

de Groot, A. D. (1977). “Memorandum: Chessinstruction in the school? A few arguments and

counterarguments,” in Chess in the Classroom.An answer to NIE, ed H. Lyman, (Saugus, MA:The Massachusetts Chess Association and theAmerican Chess Foundation), 1–10.

de Groot, A. D. (1978). Thought and Choice in Chess.The Hague: Mouton. (Original work published in1946).

Gobet, F., and Campitelli, G. (2006). “Educationand chess: a critical review,” in Chess andEducation: Selected Essays from the KoltanowskiConference, ed T. Redman (Dallas, TX: ChessProgram at the University of Texas at Dallas),124–143.

Gobet, F., de Voogt, A., and Retschitzki, J. (2004).Moves in Mind: The Psychology of Board Games.New York, NY: Psychology Press.

Hong, S., and Bart, W. (2007). Cognitive effectsof chess instruction on students at riskfor academic failure. Int. J. Spec. Educ. 22,89–96.

Kazemi, F., Yektayar, M., and Abad, A. M. B.(2012). “Investigation of the impact of chessplay on developing meta-cognitive ability andmath problem-solving power of students at dif-ferent levels ofeducation,” in 4th InternationalConference of Cognitive Science (ICCS 2011),Procedia-Social and Behavioral Sciences, Vol. 32,372–379.

Liptrap, J. M. (1998). Chess and standard test scores.Chess Life, 41–43.

Scholz, M., Niesch, H., Steffen, O., Ernst, B., Loeffler,M., Witruk, E., et al. (2008). Impact of chess train-ing on mathematics performance and concentra-tion ability of children with learning disabilities.Int. J. Spec. Educ. 23, 138–156.

Smith, J. P., and Cage, B. N. (2000). The effects ofchess instruction on the mathematics achievement

of southern, rural, black secondary students. Res.Sch. 7, 19–26.

Storey, K. (2000). Teaching beginning chess skillsto students with disabilities. Prevent. SchoolFail. Alter. Educ. Child. Youth 44, 45–40. doi:10.1080/10459880009599782

Trinchero, R. (2013). Can Chess Training Improve PisaScores in Mathematics? An Experiment in ItalianPrimary School. Paris: Kasparov Chess FoundationEurope.

Waters, A. J., Gobet, F., and Leyden, G. (2002).Visuospatial abilities of chess players. Br. J.Psychol. 93, 557–565. doi: 10.1348/000712602761381402.2002


Received: 06 March 2014; accepted: 29 June 2014;published online: 08 August 2014.Citation: Bart WM (2014) On the effect of chess train-ing on scholastic achievement. Front. Psychol. 5:762. doi:10.3389/fpsyg.2014.00762This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Bart. This is an open-accessarticle distributed under the terms of the CreativeCommons Attribution License (CC BY). The use, dis-tribution or reproduction in other forums is permit-ted, provided the original author(s) or licensor arecredited and that the original publication in thisjournal is cited, in accordance with accepted aca-demic practice. No use, distribution or reproduc-tion is permitted which does not comply with theseterms.













doi: 10.3389/fpsyg.2014.00569

Restricting range restricts conclusionsNemanja Vaci , Bartosz Gula and Merim Bilalic*

Department of General Psychology and Cognitive Science, Institute of Psychology, Alpen-Adria University Klagenfurt, Klagenfurt, Austria*Correspondence: [email protected]

Edited by:

David Zachary Hambrick, Michigan State University, USA

Reviewed by:

Fred Oswald, Rice University, USAFernand Gobet, University of Liverpool, UK

Keywords: expertise, skill acquisition, chess, Elo rating, gender differences, gerontology, talent

THE EXPERTISE APPROACH AND SKILLACQUISITIONResearch on expertise is by definitionfocused on a restricted sample of indi-viduals. Experts are people who consis-tently produce outstanding performancein their domains (Ericsson, 2006) and assuch are without exception located on thepositive side of the skill distribution. Theusual approach in the study of exper-tise is to compare the extreme group ofthe skill distribution, experts, with theextreme group at the other end, that ofnovices. This contrasting approach, whichwe have called the “expertise approach”(Bilalic et al., 2010, 2012), has a long tra-dition (Chase and Simon, 1973; Simonand Chase, 1973; De Groot, 1978; Preacheret al., 2005). Its main advantage over thecommon approach in cognition, where allparticipants are at the same skill level,is the presence of a control group ofnovices that enables falsification of resultsobtained on experts (Wason, 1960; Kuhn,1970; Campitelli and Speelman, 2013). Inthat way, the expertise approach is notunlike the neuropsychological approachthat contrasts results obtained on patientswith the results of “normal” participants(Shallice, 1988).

The main goal of the expertiseapproach is to provide evidence relatingto the cognitive and neural mechanismsbehind processes such as object and pat-tern recognition, which would be difficultto obtain from subjects who possessapproximately the same level of expertise.The skill acquisition process, which is oneof the main topics of expertise (Williamand Harter, 1899), is of secondary impor-tance in the expertise approach. This isunderstandable as the contrast between

experts and novices captures only thebeginning and the end product of theskill acquisition process. It is unrealisticto follow people for the length of timerequired in order to achieve expertisein a given domain. However, expertiseresearchers have recently started employ-ing an archival approach that provides amore complete picture of the skill acquisi-tion process (Charness and Gerchak, 1996;Chabris and Glickman, 2006; Howard,2008, 2009; Bilalic et al., 2009). In thegame of chess, a domain commonly stud-ied in expertise research, there are preciserecords of all practitioners from an earlyage (Howard, 2006a; Bilalic et al., 2009).These records include not only personalinformation such as gender and age, butalso skill levels at different stages, num-bers of games played, and correspondingskill change. The records provide a wealthof data for investigating the influence offactors such as age, gender, and even tal-ent, on the skill acquisition process. Herewe want to draw attention to the factthat some of the databases used in pre-vious research only provide records of thevery best practitioners. In the expertiseapproach, such restriction is an integralpart of the methodology, but restrictingthe range of population in the archivalapproach could have grave consequencesfor the conclusions about the nature ofskill acquisition.

DIFFERENT DATABASES, DIFFERENTCONCLUSIONSOne of the advantages of chess as a domainis that there is an objective and reliablemeasure of skill. Skill is measured on aninterval scale that reflects the performanceof a player against other players. The Elo

rating, named after Arpad Elo who intro-duced the scale as a measure of chess skill(Elo, 1978), is measured in the same wayall over the world. A beginner is supposedto have 600–800 Elo points, average play-ers about 1500 Elo points, master play-ers above 2200 Elo points, while the verybest players, called grandmasters, have rat-ings above 2500 Elo points. Expert playersare considered to have above 2000 ratingpoints.

The most frequently used database inskill acquisition studies is the databaseof the International Chess Federation,FIDE (for more information, see Howard,2006a). This database, like other chessdatabases we will mention here, offersmultiple advantages for skill acquisitionresearch. Firstly, it gathers records from the1970s to the present, and so it is possi-ble to obtain trajectories of ratings overthe course of players’ lives. Secondly, itrepresents the whole population of thevery best players in the world. Thirdly,this database contains multiple measure-ment points from players, so it can beused to observe individual skill trajec-tories. Besides rating points, numerousother variables are recorded (e.g., num-ber of games played, gender of partic-ipants, nationality, title, rating change,rating rank) which could be used forresearch purposes (Howard, 2006a). Inother words, the FIDE database offers afruitful basis for exploration and descrip-tion of the multiple factors and processesbehind chess skill acquisition.

For all its advantages, the FIDEdatabase provides only the records of thevery best players. Due to technical andlogistical reasons, the FIDE database at thebeginning logged only master level players














Vaci et al. Restricting range restricts conclusions

(above 2200 Elo). Only in the 1990s wasthe level lowered to expert level players(2000 Elo) and then in the last decade tothe level of average players (1500 Elo andbelow). In other words, the worst play-ers in the FIDE database are still averagepractitioners.

The ideal situation would be to haverecords of all players from the very begin-ning of their careers, not only when theyreach a particular level of expertise. In thisscenario, the database would also encom-pass people who for whatever reasons donot become experts. Fortunately, thereare such databases. National databases,such as the databases of the GermanChess Federation and the United StatesChess Federation (USCF), keep recordsof all their members and thus repre-sent the whole population of competi-tive (national) players. They provide allthe information the FIDE database offerswithout restricting the range of skill.

If one is trying to examine factors thatinfluence skill acquisition, a database thatcontains only average and above-averageplayers should not be the starting point

of investigations. As is well known inthe field of statistics, the conclusionsobtained on data with restricted rangecould be misleading especially if theyare not obtained by appropriate analy-ses (Long, 1997; Sackett and Yang, 2000).The best possible way to examine dif-ferences in effects that researchers couldobtain while analyzing the data is throughthe comparison of effects made on differ-ent databases. Here we illustrate the pos-sible pitfalls of skill range restriction bycomparing distributions of ratings froma database with restricted range (FIDE)and a database with unrestricted range(German).

The FIDE and German databases con-tain similar number of practitioners,around 120,000 (see Figure 1). However,the ratings of the two distributions overlaponly at the highest values of the Germandistribution and the lowest values of FIDEdistribution. Not only are the mean andvariance of both distribution vastly differ-ent, but also other parts of the distribu-tion, such as quartiles, are also extremelydifferent.

Restricting the database to the bestplayers also has a consequence for the skilltrajectories of players. One needs time tobecome an expert and the players in theFIDE database are older (37 years) than theplayers in the German database (32 years).The differences between the two databasesare also evident when we compare typi-cal skill trajectories. Figure 2A shows theFIDE players entering the database ataround age 10, having already becomecompetent players (rating of 1900 Elo),with a subsequent shallow increase to thepeak at age 39. In contrast, the Germanplayers have a steeper increase, since theyare entered the database as novices andlearn faster at beginning skill level, asimplied by the power law of practice(Newell and Rosenbloom, 1981), until thesame peak at age 39. The decline in lateryears is also different in the two databaseswith FIDE players declining faster thantheir German counterparts.

One could say that the FIDE andGerman players have vastly different rat-ings that make the comparison betweenthem difficult. One way around this

FIGURE 1 | Distribution of chess skill as measured by Elo rating in

FIDE (blue color) and German (red) databases (as of 2008). Thedatabases contain similar number of players, but differ vastly in the

distribution shape and coverage—the only overlap is at the highestvalues of the German database and lowest values of the FIDEdatabase.






FIGURE 2 | (A) Average skill trajectories, in Elo with 95% confidence intervals, over the years in FIDE database (blue color) and German (red) database. (B)

Average standardized skill trajectories, in Z -values with 95% confidence intervals, over the years in FIDE database (blue color) and German (red) database.

problem is to standardize ratings in eachdatabase separately and check if the skilltrajectories are similar in both datasets.Figure 2B shows the standardized ratingas a function of age. Again, on aver-age FIDE players start at a higher skilllevel but improve more slowly. They alsohave a higher peak but their decline isso rapid that in the latter stages of theircareers their standardized performance islower than the standardized performanceof their German colleagues.

EXPLAINING CONTRADICTORYRESULTSWe have demonstrated that there are vastdifferences between the databases com-monly used in the archival approach toskill acquisition. The two datasets are com-pletely different in the range of valuesas well as in the number of participantsthat are obtaining a particular rating. Therestricted databases, such as FIDE, do notrepresent the whole skill range and mayprovide inadequate answers to the ques-tions under investigation. The restrictionof range and its consequences may alsoexplain some of the inconsistencies andcontradictory findings in the field.

For example, Roring and Charness(2007) used the FIDE databases to inves-tigate age effects on skill acquisition. Theydemonstrated the peak age in chess skill

to be around 43 years, much later thanpreviously proposed peak around 35 years(e.g., Howard, 2005). Another surpris-ing result was the fact that the declineis steeper for initially lower rated par-ticipants than for higher rated partici-pants. In other words, initially more ableparticipants were declining significantlymore slowly than their initially weaker col-leagues. Our illustrations (Figures 2A,B)indicate that both peak age and decliningrate are influenced by the range restric-tion in the FIDE database. The conclu-sion would be significantly different if awhole range database, such as the Germandatabase, were used.

To further illustrate possible conse-quences of the range restriction, wecan consider the inconsistent findingsin the research on gender differencesin skill acquisition. It is notable thatthe studies using the restricted FIDEdatabase regularly find gender differencesin skill acquisition (Howard, 2005, 2006b,2014 but see, Bilalic and McLeod, 2006,2007). Furthermore, the studies using thenational German and USCF databases(Chabris and Glickman, 2006; Bilalic et al.,2009) also noted the differences in themean and highest ratings of women andmen. However, using the unrestrictedrange and the full lifespan data, theyobserved that factors such as participation

rates and dropout rates could explainthe differences. This kind of analysis isimpossible with the FIDE database wheredropouts are not recorded because thepeople concerned stopped playing chessbefore they achieved expert level.

Researchers using the FIDE databaseto investigate talent (Howard, 2008, 2009)face a similar problem. The time requiredto reach a certain level and the amountof practice (as measured by games played)may well provide clues about the differ-ent natural endowments of certain players.This in turn may allow us to speculateabout different levels of talent. It is, how-ever, impossible to make any certain con-clusions if we lack the very first part oftheir skill acquisition process, as we doin the FIDE database. As with the gen-der factor in skill acquisition process, thedifferences in the early stages may as wellovershadow the differences at the highestlevel. Similarly, the causes behind dropoutsmay remain unresolved because the dataof the people who for whatever reasonsstopped playing chess is not available.

Both the expertise and archivalapproaches are important vehicles for theinvestigation of expertise and cognition ingeneral. The restricted range of focus inthe expertise approach is a fundamentalpart of the methodology and an advan-tage over usual research on cognition. The





archival approach offers the possibility ofcapturing the full cycle of the long-termskill acquisition process. In this paperwe have demonstrated that the resultsobtained on the restricted range do notnecessarily generalize to the whole range ofvalues. The effects obtained with restrictedrange cannot and should not be used tomake inferences about the mechanismsand factors that influence skill acquisition.When we restrict our data, we restrict ourconclusions too.

ACKNOWLEDGMENTWe would like to thank Frank Hoppefor providing the German database andRobert Gaschler for the FIDE database.

REFERENCESBilalic, M., Langner, R., Erb, M., and Grodd, W.

(2010). Mechanisms and neural basis of objectand pattern recognition: a study with chessexperts. J. Exp. Psychol. Gen. 139, 728–742. doi:10.1037/a0020756

Bilalic, M., and McLeod, P. (2006). How intellectualis chess?—a reply to Howard. J. Biosoc. Sci. 38,419–421. doi: 10.1017/S0021932005001185

Bilalic, M., and McLeod, P. (2007). Participation ratesand the difference in performance of women andmen in chess. J. Biosoc. Sci. 39, 789–793. doi:10.1017/S0021932007001861

Bilalic, M., Smallbone, K., McLeod, P., and Gobet,F. (2009). Why are (the best) women so goodat chess? participation rates and gender differ-ences in intellectual domains. Proc. Biol. Sci. 276,1161–1165. doi: 10.1098/rspb.2008.1576

Bilalic, M., Turella, L., Campitelli, G., Erb, M., andGrodd, W. (2012). Expertise modulates the neu-ral basis of context dependent recognition ofobjects and their relations. Hum. Brain Mapp. 33,2728–2740. doi: 10.1002/hbm.21396

Campitelli, G., and Speelman, C. (2013). Expertiseparadigms for investigating the neural substratesof stable memories. Front. Hum. Neurosci. 7:740.doi: 10.3389/fnhum.2013.00740

Chabris, C. F., and Glickman, M. E. (2006). Sexdifferences in intellectual performance: analysis

of a large cohort of competitive chess players.Psychol. Sci. 17, 1040–1046. doi: 10.1111/j.1467-9280.2006.01828.x

Charness, N., and Gerchak, Y. (1996). Participationrates and maximal performance: a log-linearexplanation for group differences, such as Russianand male dominance in chess. Psychol. Sci. 7,46–51. doi: 10.1111/j.1467-9280.1996.tb00665.x

Chase, W. G., and Simon, H. A. (1973). Perception inchess. Cogn. Psychol. 4, 55–81.

De Groot, A. (1978). Thought and Choice in Chess, 2ndEdn. The Hague: Mouton De Gruyter.

Elo, A. E. (1978). The Rating of Chessplayers, Pastand Present, Vol. 3. London: B.T. Batsford Ltd.Available online at: http://www.getcited.org/pub/101876597

Ericsson, K. A. (2006). The Cambridge Handbook ofExpertise and Expert Performance. New York, NY:Cambridge University Press.

Howard, R. W. (2005). Objective evidence of risingpopulation ability: a detailed examination of lon-gitudinal chess data. Pers. Individ. Dif. 38, 347–363.doi: 10.1016/j.paid.2004.04.013

Howard, R. W. (2006a). A complete database ofinternational chess players and chess performanceratings for varied longitudinal studies. Behav.Res. Methods 38, 698–703. doi: 10.3758/BF03193903

Howard, R. W. (2006b). IQ, visuospatial abil-ity and the gender divide: a reply to Bilalicand Mcleod. J. Biosoc. Sci. 38, 423–426. doi:10.1017/S0021932005001197

Howard, R. W. (2008). Linking extreme precocity andadult eminence: a study of eight prodigies at inter-national chess. High Abil. Stud. 19, 117–130. doi:10.1080/13598130802503991

Howard, R. W. (2009). Individual differences in exper-tise development over decades in a complex intel-lectual domain. Mem. Cognit. 37, 194–209. doi:10.3758/MC.37.2.194

Howard, R. W. (2014). Gender differences in intel-lectual performance persist at the limits of indi-vidual capabilities. J. Biosoc. Sci. 46, 386–404. doi:10.1017/S0021932013000205

Kuhn, T. S. (1970). The Structure of ScientificRevolutions. Chicago, IL: University of ChicagoPress.

Long, J. S. (1997). Regression Models for Categoricaland Limited Dependent Variables, Vol. 7. ThousandOaks, CA: Sage.

Newell, A., and Rosenbloom, P. S. (1981).“Mechanisms of skill acquisition and the law of

practice,” in Cognitive Skills and Their Acquisition,ed J. R. Anderson (Hillsdale, NJ: Erlbaum), 1–55.

Preacher, K. J., Rucker, D. D., MacCallum, R. C., andNicewander, W. A. (2005). Use of the extremegroups approach: a critical reexamination and newrecommendations. Psychol. Methods 10, 178–192.doi: 10.1037/1082-989X.10.2.178

Roring, R. W., and Charness, N. (2007). A multi-level model analysis of expertise in chess acrossthe life span. Psychol. Aging 22, 291–299. doi:10.1037/0882-7974.22.2.291

Sackett, P. R., and Yang, H. (2000). Correctionfor range restriction: an expanded typology.J. Appl. Psychol. 85, 112–118. doi: 10.1037/0021-9010.85.1.112

Shallice, T. (1988). From Neuropsychology to MentalStructure. New York, NY: Cambridge UniversityPress.

Simon, H. A., and Chase, W. G. (1973). Skill inchess: experiments with chess-playing tasks andcomputer simulation of skilled performance throwlight on some human perceptual and memoryprocesses. Am. Sci. 61, 394–403.

Wason, P. C. (1960). On the failure to eliminatehypotheses in a conceptual task. Q. J. Exp. Psychol.12, 129–140. doi: 10.1080/17470216008416717

William, L. B., and Harter, N. (1899). Studies onthe telegraphic language: the acquisition of ahierarchy of habits. Psychol. Rev. 6, 345–375. doi:10.1037/h0073117


Received: 05 April 2014; accepted: 22 May 2014;published online: 12 June 2014.Citation: Vaci N, Gula B and Bilalic M (2014)Restricting range restricts conclusions. Front. Psychol.5:569. doi: 10.3389/fpsyg.2014.00569This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Vaci, Gula and Bilalic. This is anopen-access article distributed under the terms of theCreative Commons Attribution License (CC BY). Theuse, distribution or reproduction in other forums is per-mitted, provided the original author(s) or licensor arecredited and that the original publication in this journalis cited, in accordance with accepted academic practice.No use, distribution or reproduction is permitted whichdoes not comply with these terms.


http://www.getcited.org/pub/101876597

http://www.getcited.org/pub/101876597












OPINION ARTICLEpublished: 09 July 2014

doi: 10.3389/fpsyg.2014.00707

A proposed integration of the expert performance andindividual differences approaches to the study of eliteperformanceScott Barry Kaufman1,2*

1 The Imagination Institute, Philadelphia, PA, USA2 Positive Psychology Center, University of Pennsylvania, Philadelphia, PA, USA*Correspondence: [email protected]

Edited by:

David Zachary Hambrick, Michigan State University, USA

Reviewed by:

Fernand Gobet, University of Liverpool, UK

Keywords: expertise, intelligence, motivation, individual differences, expert performance, inspiration, general cognitive ability

I recently had the pleasure of editinga volume of essays on the determinantsof greatness (Kaufman, 2013a). A vari-ety of perspectives were represented inthe volume, including behavioral genetics,individual differences, and expert perfor-mance. The clearest conclusion from thevolume was that the development of highachievement involves a complex interac-tion of many personal and environmentalvariables that feed off each other in non-linear, mutually reinforcing, and nuancedways, and that the most complete under-standing of the development of elite per-formance can only be arrived through anintegration of perspectives.

To help spur more integration, I sug-gest that cognitive psychologists who arestudying deliberate practice and chunking,and individual differences researchers whoare investigating cognitive ability and per-sonality, focus more on common ground.I’ve noticed that the debate often ends upbeing “innate talent vs. deliberate prac-tice” (see Ericsson et al., 2007; Ericsson,2014), when that false dichotomy is detri-mental to scientific progress (Gobet, 2013;Kaufman, 2013a). Deliberate practice—defined by Ericsson (2013) as “engage-ment with full concentration in a trainingactivity designed to improve a particu-lar aspect of performance with immediatefeedback, [and] opportunities for grad-ual refinement by repetition and problemsolving”—depends on many traits whichvary in the general population, and whichhave a genetic basis. But that doesn’tmean that heritable traits are necessarily“immutable constraints on the acquisition

of various types of expert performance”(Ericsson, 2014).

Given our current state of scientificknowledge, I hope we can all agree that:

• There is no such thing as “innate tal-ent.” All skills require practice and sup-port for their development (Kaufman,2013b).

• The sheer number of hours engagedin practice is not as important as thequality of deliberate practice (Ericsson,2013).

• There is nothing magical about 10,000 hof deliberate practice: the average hoursof deliberate practice associated withexpert performance varies by domain,and within domains, varies amongindividuals (Ericsson, 2013; Kaufman,2013a).

• Deliberate practice does not explain allof the variation in elite performance(Ericsson, 2013; Hambrick et al., 2014).

• Other traits beyond deliberate prac-tice are critical for the development ofexpert performance.

• Virtually all psychological traits areinfluenced by a complex, dynamic inter-play of genetic and environmental fac-tors (Johnson et al., 2009).

• Individual differences at any singlemoment of time don’t necessarily con-strain ultimate levels of performance,even though they may influence the rateof expertise acquisition.

Assuming researchers can agree on theseseven basic principles, a fruitful researchdirection is the investigation of the manner

in which individual differences influence(but not necessarily constrain) the devel-opment of expertise. One mode of oper-ation is by influencing the efficiency ofexpertise acquisition, therefore speedingup the rate of acquisition. Ericsson (2013)acknowledges that the 10,000 h of prac-tice he found among elite violinists atage 20 was just an average, with substan-tial variation around the mean. In fact,Simonton has found across the arts, sci-ences, and leadership, that those with thegreatest lifetime productivity and high-est levels of eminence required the leastamount of time to acquire the requisiteexpertise (Simonton, 1991a,b, 1992, 1997,1999).

General cognitive ability is one fac-tor that can influence the efficiency ofexpertise acquisition. Individual differ-ences researchers have spent over 100 yearsstudying patterns of variation in cognitiveability (e.g., Carroll, 1993; Jensen, 1998).Brain imaging studies support the ideathat people who do well on tests of cog-nitive ability use fewer brain resources tosolve novel and complex problems (Haieret al., 1992; Neubauer and Fink, 2009; Vanden Heuvel et al., 2009; Deary et al., 2010;Prabhakaran et al., 2011). Unfortunately,this literature (which emphasizes cogni-tive efficiency) is not well integrated withthe research of cognitive psychologists whoemphasize deliberate practice, chunking,and strategy use. However, I believe thesevarious approaches are better suited forintegration than it may seem at first blush.

Consider a set of studies conductedby Bor and colleagues, in which they












Kaufman A proposed integration

found that chunking consistently activatesthe prefrontal-parietal brain network (Boret al., 2004; Bor and Owen, 2007; Bor,2012; Bor and Seth, 2012). Bor and Owen(2007) had participants memorize unfa-miliar verbal and numerical double-digitsequences. The sequences were either ran-domly arranged (e.g., 31, 24, 89, 65)—and therefore not conducive to the use ofstrategies—or structured (e.g., 57, 68, 79,90)—which made them amenable to theuse of chunking strategies. The prefrontal-parietal brain network was consistentlymost active during the structured trials,even though the unstructured trials placeda higher demand on working memory,and were more difficult for participants tomemorize.

The prefrontal-parietal network hasalso been heavily implicated on tests ofworking memory and general cognitiveability (Prabhakaran et al., 2000; Jungand Haier, 2007; Colom et al., 2009).The research of Bor and colleagues sug-gests that one of the primary functionsof the prefrontal-parietal brain networkis the conscious detection of patterns,which aids in the efficiency of learning.Indeed, Spearman (1904) argued that thebest measure of general cognitive abilityrequires grasping relationships, inferringrules, noticing similarities and differences,and “educing” (Lating for “drawing out”)the relevant relations in a complex pattern.Indeed, the Ravens Progressive Matricestest—which is strongly correlated with thegeneral cognitive ability factor—appearsto measure these skills (Conway et al.,2003). The Ravens test places a heavyburden on working memory because youmust engage in fluid reasoning on thespot, with no external aids and often withstrict time limits. However, those who havemore efficient cognitive strategies for less-ening the cognitive load will be at a distinctadvantage in this testing environment.

Consistent with this idea, Nandagopalet al. (2010) had twins think aloudwhile they solved various tasks, includ-ing an associative learning task that issignificantly correlated with general cog-nitive ability (see Kaufman et al., 2009).They found that performance on testsof cognitive ability were heavily influ-enced by the use of strategies, anddifferences in strategy use on an associa-tive learning task (which was amenable to

use of strategies) explained a significantamount of the genetic influences onperformance.

Their study raises the intriguing sug-gestion that the heritability of generalcognitive ability may be due, in part, tothe ability to efficiently chunk informa-tion in working memory. Therefore, whileEricsson (2014) may be right that cogni-tive ability does not necessarily constrainthe acquisition of expertise, it’s still entirelypossible that cognitive ability influences theefficiency and rate of expertise acquisi-tion (especially when expertise acquisitiondraws heavily on general cognitive ability;2014 special issue). Consistent with this,Meinz and Hambrick (2010) found thatalthough deliberate practice accountedfor 45.1% of the variation in pianosight-reading performance among expertpianists, working memory accounted foran additional 7.4% of the variance.

Of course, cognitive efficiency isn’t theonly way that individual differences caninfluence expertise acquisition. Anothermode of operation is by sustaining themotivation to practice over an extendedperiod of time. Ericsson et al. (1993)acknowledged this possibility when theysay: “It is quite plausible, however, thatheritable individual differences mightinfluence processes related to motiva-tion and the original enjoyment of theactivities in the domain and, even moreimportant, affect the inevitable differ-ences in the capacity to engage in hardwork (deliberate practice)” (p. 399). EvenArthur Jensen (one of the biggest propo-nents of general cognitive ability) onceconcluded that “some kind of motiva-tional factor that sustains enormous andprolonged interest and practice in a par-ticular skill probably plays a larger partin extremely exceptional performance thandoes psychometric g or the speed of ele-mentary information processes (Jensen,1990, p. 259, italics added).”

I believe an overlooked characteris-tic that influences the motivation toengage in deliberate practice is inspi-ration (Kaufman, 2013b). When peoplebecome inspired, they usually are inspiredto realize some future image of them-selves (Torrance, 1983). It is the clarity ofthis vision, and the belief that the visionis attainable, that can propel a personfrom apathy to engagement, and sustain

the energy to engage in deliberate practiceover the long haul, despite obstacles andsetbacks. Indeed, Todd Thrash, AndrewElliot, and colleagues have conductedmultiple studies showing that inspiration(measured both as a trait and a motiva-tional state) is associated with an approachmotivation, positive emotions, and anincrease in creative productivity (Thrashand Elliot, 2003, 2004; Thrash et al., 2010).

In fact, in one of their studies (Thrashet al., 2010), inspiration not only predictedthe creativity of writing samples in scienceand poetry, but also increased the efficiencyof the writing samples (e.g., a larger num-ber of typed words that were retained inthe final product, and less time pausingand more time writing). This raises theintriguing idea that motivational charac-teristics may cause an increase in cognitiveefficiency, which would ultimately increasethe rate of expertise acquisition. I believethis is a promising area for future research.

These are just a few examples ofhow the cognitive psychology approach toexpertise and the investigation of individ-ual differences can be more tightly inte-grated. To conclude: while others have sug-gested the importance of computer mod-eling for integration (Gobet, 2013), I haveargued here that other important con-tributors to scientific progress are accu-rate framing of the issues, standing ona common ground of assumptions, andinvestigating the influence of traits on thedevelopment of expertise.

ACKNOWLEDGMENTThanks to Zach Hambrick for his encour-agement to submit this paper.

REFERENCES(2014). Acquiring expertise: ability, practice, and

other influences. Intelligence 45, 1–124 (specialissue).

Bor, D. (2012). The Ravenous Brain: How the NewScience of Consciousness Explains Our InsatiableSearch for Meaning. New York, NY: Basic Books.

Bor, D., Cumming, C. N., Scott, E., and Owen,A. M. (2004). Prefrontal cortical isnvolvementin verbal encoding strategies. Eur. J. Neurosci.18, 3365–3370. doi: 10.1111/j.1460-9568.2004.03438.x

Bor, D., and Owen, A. M. (2007). A commonprefrontal-parietal network for mnemonic andmathematical recoding strategies within work-ing memory. Cereb. Cortex 17, 778–786. doi:10.1093/cercor/bhk035

Bor, D., and Seth, A. K. (2012). Consciousness andthe prefrontal parietal network: insights from





Kaufman A proposed integration

attention, working memory, and chunking. Front.Psychol. 3:63. doi: 10.3389/fpsyg.2012.00063

Carroll, J. B. (1993). Human Cognitive Abilities.Cambridge: Cambridge University Press.

Colom, R., Haier, R. J., Head, K., Alvarez-Linera, J.,Quiroga, M. A., Shih, P. C., et al. (2009). Graymatter correlates of fluid, crystallized, and spatialintelligence: testing the P-FIT model. Intelligence37, 124–135. doi: 10.1016/j.intell.2008.07.007

Conway, A. R. A., Kane, M. J., and Engle, R. W. (2003).Working memory capacity and its relation to gen-eral intelligence. Trends Cogn. Sci. 7, 547–552. doi:10.1016/j.tics.2003.10.005

Deary, I. J., Penke, L., and Johnson, W. (2010).The neuroscience of human intelligence dif-ferences. Nat. Rev. Neurosci. 11, 201–211. doi:10.1038/nrn2793

Ericsson, K. A. (2013). Training history, deliberatepractice and elite sports performance: an analysisin response to Tucker and Collins review—whatmakes champions? Br. J. Sports Med. 47, 533–535.doi: 10.1136/bjsports-2012-091767

Ericsson, K. A. (2014). Why expert performance isspecial and cannot be extrapolated from stud-ies of performance in the general population: aresponse to criticisms. Intelligence 45, 81–103. doi:10.1016/j.intell.2013.12.001

Ericsson, K. A., Krampe, R. T., and Tesch-Römer,C. (1993). The role of deliberate practice inthe acquisition of expert performance. Psychol.Rev. 100, 363–406. doi: 10.1037/0033-295X.100.3.363

Ericsson, K. A., Roring, R. W., and Nandagopal, K.(2007). Giftedness and evidence for reproduciblysuperior performance: an account based on theexpert performance framework. High Abil. Stud.18, 3–56. doi: 10.1080/13598130701350593

Gobet, F. (2013). Expertise vs. talent. Talent Dev.Excellence 5, 59–70. Available online at: http://www.iratde.org/journal/issues/112-issue-20131

Haier, R. J., Siegel, B., Tang, C., Abel, L., andBuschbaum, M. S. (1992). Intelligence and changesin regional cerebral glucose metabolic-rate fol-lowing learning. Intelligence 16, 425–426. doi:10.1016/0160-2896(92)90018-M

Hambrick, D. Z., Oswald, F. L., Altmann, E. M.,Meinz, E. J., Gobet, F., and Campitelli, G. (2014).Deliberate practice: is that all it takes to become anexpert? Intelligence 45, 34–45. doi: 10.1016/j.intell.2013.04.001

Jensen, A. (1998). The g Factor. New York, NY: Praeger.Jensen, A. R. (1990). Speed of information processing

in a calculating prodigy. Intelligence 14, 259–274.doi: 10.1016/0160-2896(90)90019-P

Johnson, W., Turkheimer, E., Gottesman, I. I., andBouchard, T. J. Jr. (2009). Beyond heritability:twin studies in behavioral research. Curr. Dir.Psychol. Sci. 18, 217–220. doi: 10.1111/j.1467-8721.2009.01639.x

Jung, R. E., and Haier, R. J. (2007). The parieto-frontalintegration theory (P-FIT) of intelligence: con-verging neuroimaging evidence. Behav. Brain Sci.30, 135–154. doi: 10.1017/S0140525X07001185

Kaufman, S. B. (2013a). The Complex of Greatness:Beyond Talent or Practice. New York, NY: OxfordUniversity Press.

Kaufman, S. B. (2013b). Ungifted: IntelligenceRedefined. New York, NY: Basic Books.

Kaufman, S. B., DeYoung, C. G., Gray, J. R., Brown,J., and Mackintosh, N. (2009). Associative learn-ing predicts intelligence above and beyondworking memory and processing speed.Intelligence 37, 374–382. doi: 10.1016/j.intell.2009.03.004

Meinz, E. J., and Hambrick, D. Z. (2010). Deliberatepractice is necessary but not sufficient toexplain individual differences in piano sight-reading skill. Psychol. Sci. 21, 914–919. doi:10.1177/0956797610373933

Nandagopal, K., Roring, R., Ericsson, K. A., andTaylor, J. (2010). Strategies may mediate herita-ble aspects of memory performance: a twin study.Cogn. Behav. Neurol. 23, 224–230. doi: 10.1097/WNN.0b013e3181e07d29

Neubauer, A. C., and Fink, A. (2009). Intelligenceand neural efficiency. Neurosci. Biobehav. Rev. 33,1004–1023. doi: 10.1016/j.neubiorev.2009.04.001

Prabhakaran, V., Narayanan, K., Zhao, Z., andGabrieli, J. D. (2000). Integration of diverse infor-mation in working memory within the frontallobe. Nat. Neurosci. 3, 85–90. doi: 10.1038/71156

Prabhakaran, V., Rypma, B., Narayanan, N. S., Meier,T. B., Austin, B. P., Nair, V. A., et al. (2011).Capacity-speed relationships in prefrontal cortex.PLoS ONE 6:e27504. doi: 10.1371/journal.pone.0027504

Simonton, D. K. (1991a). Career landmarks inscience: individual differences and interdisci-plinary contrasts. Dev. Psychol. 27, 119–130. doi:10.1037/0012-1649.27.1.119

Simonton, D. K. (1991b). Emergence and realizationof genius: the lives and works of 120 classicalcomposers. J. Pers. Soc. Psychol. 61, 829–840.

Simonton, D. K. (1992). Leaders of American psy-chology, 1879–1967: career development, creativeoutput, and professional achievement. J. Pers. Soc.Psychol. 62, 5–17. doi: 10.1037/0022-3514.62.1.5

Simonton, D. K. (1997). Creative productivity: a pre-dictive and explanatory model of career trajecto-ries and landmarks. Psychol. Rev. 104, 66–89. doi:10.1037/0033-295X.104.1.66

Simonton, D. K. (1999). Talent and its development:an emergenic and epigenetic model. Psychol.Rev. 106, 435–457. doi: 10.1037/0033-295X.106.3.435

Spearman, C. (1904). General intelligence: objectivelydetermined and measured. Am. J. Psychol. 15,201–293. doi: 10.2307/1412107

Thrash, T. M., and Elliot, A. J. (2003). Inspiration asa psychological construct. J. Pers. Soc. Psychol. 84,871–889. doi: 10.1037/0022-3514.84.4.871

Thrash, T. M., and Elliot, A. J. (2004). Inspiration: corecharacteristics, component processes, antecedents,and function. J. Pers. Soc. Psychol. 87, 957–973. doi:10.1037/0022-3514.87.6.957

Thrash, T. M., Maruskin, L. A., Cassidy, S. E., Fryer,J. W., and Ryan, R. M. (2010). Meditating betweenthe muse and the masses: inspiration and the actu-alization of creative ideas. J. Pers. Soc. Psychol. 98,469–487. doi: 10.1037/a0017907

Torrance, E. P. (1983). The importance of falling inlove with “something.” Creativ. Child Adult Q. 8,72–78.

Van den Heuvel, M. P., Stam, C. J., Kahn, R. S.,and Hulshoff Pol, H. E. (2009). Efficiency offunctional brain networks and intellectual perfor-mance. J. Neurosci. 29, 7619–7624. doi: 10.1523/JNEUROSCI.1443-09.2009


Received: 31 May 2014; accepted: 19 June 2014;published online: 09 July 2014.Citation: Kaufman SB (2014) A proposed integrationof the expert performance and individual differencesapproaches to the study of elite performance. Front.Psychol. 5:707. doi: 10.3389/fpsyg.2014.00707This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Kaufman. This is an open-accessarticle distributed under the terms of the CreativeCommons Attribution License (CC BY). The use, dis-tribution or reproduction in other forums is permitted,provided the original author(s) or licensor are creditedand that the original publication in this journal is cited,in accordance with accepted academic practice. No use,distribution or reproduction is permitted which does notcomply with these terms.


http://www.iratde.org/journal/issues/112-issue-20131

http://www.iratde.org/journal/issues/112-issue-20131











OPINION ARTICLEpublished: 04 March 2014

doi: 10.3389/fpsyg.2014.00186PSYCHOLOGY

Expertise: defined, described, explainedLyle E. Bourne Jr.1*, James A. Kole2 and Alice F. Healy1

1 Department of Psychology and Neuroscience, University of Colorado, Boulder, CO, USA2 School of Psychological Sciences, University of Northern Colorado, Greeley, CO, USA*Correspondence: [email protected]

Edited by:

David Z. Hambrick, Michigan State University, USA

Reviewed by:

David Z. Hambrick, Michigan State University, USAJonathan Wai, Duke University, USA

Keywords: expertise, definition, description, explanation, mind, brain

INTRODUCTIONScience aims to define, describe, andexplain significant natural phenomena.Each of these goals of science suggests anincreasingly deeper understanding of thetarget phenomenon. We discuss in thispaper how these goals are or might berealized in the science of expertise.

DEFINITIONDefinitions are given in an attempt toidentify phenomena and to delineateexamples from non-examples. Expertiseis consensually defined as elite, peak, orexceptionally high levels of performanceon a particular task or within a givendomain. One who achieves this status iscalled an expert or some related term,such as virtuoso, master, maven, prodigy,or genius. These terms are meant to labelsomeone whose performance is at the topof the game. An expert’s field of exper-tise can be almost anything from crafts-manship, through sports and music, toscience or mathematics. People usuallyagree on examples of expertise, like Yo-YoMa (musical performance), Fred Astaireand Ginger Rogers (ballroom dancing),Antiques Roadshow Appraisers, AlbertEinstein (physics), Tiger Woods (golf),Bette Davis (acting), Nelson Mandela(politics), or Hillary Rodham Clinton(international relations).

Why different terms? Each term car-ries with it a slightly nuanced meaning.Shaded meanings vary in their empha-sis on experience or constitutional factorsas the source of high levels of perfor-mance. The term chosen to characterizesuperior performance carries with it animplied cause. Like expert, virtuoso ormaster is the result of hard work and long

training. If talent is involved, it is a tal-ent for hard labor. In contrast, prodigy, likegenius, results from an endowment, whichshows up early in life without the benefitof training.

It might be appealing to the laypersonto believe that a genius is just born thatway. Elite performance just comes natu-ral to a genius; you don’t have to invest allthat time and effort on training, because ifyou don’t have what it takes you’ll never getthere. Moreover, you don’t have to explainwhy you never had a significant insight,because you just didn’t inherit the rightabilities or genes. But the facts seem to bethat, although people do differ in some-thing called ability or talent, in sports ormedicine or any area of human endeavor,talent is a necessary starting point, a plat-form from which to begin. To become anelite performer one has to capitalize on hisor her abilities. Training is the sine qua non.

Consider a specific case. Pablo Picasso,Spanish painter and sculptor, was one ofthe greatest and most influential artistsof the 20th century. Born into a familythat cultivated the arts, he demonstratedextraordinary artistic ability at an earlyage, encouraged by his parents. All the ele-ments were in place for Picasso—paints,brushes, canvases, and parents who couldrecognize good artistic work. Paintingin the beginning in a naturalistic man-ner, his style changed later in life as heexperimented with different theories, tech-niques, and ideas, for example, creating(with Braque) a unique style that has cometo be known as cubism. There is no doubtthat Picasso was a child prodigy. He hadan ability to create significant objects thatthe art world and collectors recognizedearly on for their value. He seems to have

been endowed with pure genius for paint-ing and sculpting. But it is less often rec-ognized that he was trained classically inthe arts and that he worked incessantlyat his craft, devoting long hours day andnight. And, over time, the quality of hiswork improved, as judged by his peers,and expanded into previously unexploredareas and techniques. He could producenew paintings later in life quickly, someconsisting of little more than three or fourstrokes of his pen, and more or less at will,each of them a virtuoso performance. Butthat performance was based on a level ofexpertise achieved, by dint of hard work, byfew other mortals. Picasso is but one caseof expertise and, as such, cannot validate ageneral rule. Nonetheless, his accomplish-ments are clearly based on a combinationof ability and effort, a characteristic thatother experts share.

DESCRIPTIONWe all know an expert when we see one.Normally people will quickly recognize thedifference between expertise and normalor ordinary performance in any domain.Expertise, itself, is a descriptive term. Todescribe is to add detail in the specific caseto a more general definition. A descrip-tion of expertise requires an inventory ofwhat the expert knows, knows how todo, wants or intends to do, and what heor she does or achieves. Psychologically,knowledge and skills are mental or cogni-tive concepts. They are not material enti-ties, known by their physical make-up,but rather they are states of mind. Thisfact alone does not make them unscien-tific. Rather they are quite sound scien-tific concepts, known by their function, bythe behavior potential they provide. Mind,

www.frontiersin.org March 2014 | Volume 5 | Article 186 | 211












Bourne et al. Expertise: defined, described, explained

knowledge, skill, and other cognitive con-cepts are analogous to gravity in physics orevolution in biology, understood in termsof their effects or functions, not by theirmaterial structure.

Obviously, there is more to exper-tise that just acquiring the right knowl-edge and skills. Expertise is based insome measure on the resources a personcomes equipped with, his or her natu-ral talent or biological endowment. Weput an emphasis on practice and expe-rience primarily because their contribu-tion to expert performance is too oftenoverlooked or minimized by the layperson(Ericsson et al., 1993). But clearly, inher-ited prodigiousness, body characteristics,dexterity, and the like, which are part andparcel of the equipment we come to anytask with, all play a role, allowing somepeople just to be flat out better preparedthan others. These natural factors provideessentially a foundation for expertise inany task. Given abilities and potential areuseless unless they are capitalized on oractivated by experience and practice, and,conversely, practice might be futile if onedoesn’t have some initial capacity. Bothendowment and experience must be a partof a complete description of expertise.

Thus, to describe expertise is to identifythe endowed resources, catalog the knowl-edge, and specify the skills of a person whois capable of performing in some domainat the very highest level, achieved by fewothers (perhaps by only a very small per-centage of the general population).

EXPLANATIONHow do experts get to be experts? What’sthe explanation? Is there something deeperthan a description that we need to knowabout expertise? Maybe the brain or genesare at the bottom of it.

Mind and brain are often conflatedterms and used interchangeably (Bourneand Healy, 2014). It is tempting to equatemind and brain, and it’s quite commonlydone among psychologists and other sci-entists, not to mention laypersons. Brainscanning is often used to study “howthe mind works.” The general assump-tion behind brain-scanning procedures isthat the brain provides a mechanism formental functions. Thus, you commonlycome across phrases such as, “How totrain the brain,” “The brain learns (this,

that, or the other thing),” “Learning is arewiring of the brain,” or “The brain is themind’s machine.” The implication is thatthe brain causes thinking and behavior tobe what they are. The psychological aspectsof behavior are caused by a material, bio-logical entity called the brain. Thus, theultimate explanation for why and how webehave as we do is to be found in a mate-rial thing called the brain. In theories ofthis type, the brain is the deus ex machinethat resolves difficulties we might havein understanding why people behave asthey do.

But the facts of the matter are dif-ferent. Training, experience, and practicedirectly change the mind (i.e., thought andbehavior), but only indirectly the brain.It is a person or a mind, defined by thecollection of all current knowledge andskill, that is trained, not a brain. A per-son learns, not the brain. That is notto say that the brain and what goes onin the brain are irrelevant, inconsequen-tial, or unimportant in skill or knowl-edge acquisition. Quite the contrary, whathappens in the brain as we learn andbehave is essential to understanding themind. As thinking happens, so do brainprocesses. Mind and brain processes aretime-locked, and one can actually mea-sure brain changes during thought. Still,there is no good reason to believe thatone of these processes, say, brain activ-ity, is more fundamental or causes theother, thought or behavior, to be what itis. In fact, the other causal direction—that thought causes brain activity to bewhat it is—is just as plausible. Considerthe possibility that neither causes the otherin a direct way but that both are goingon in parallel simultaneously and in aninterrelated way at all times. We thinkof that position as consistent with thelong accepted first principle of the unityof the sciences. What we observe to betrue in one domain of science shouldnot conflict with what we observe at thesame time in another domain of science.What we observe to be true psychologi-cally should be consistent with what weobserve biologically (or chemically, phys-ically, etc.). Thus, mind (psychological)and brain (biological) are unique but dif-ferent, and both will reflect, in their dif-ferent ways, the expression of expertise inbehavior.

So what is the explanation for exper-tise? Consider this, can you explain some-thing you cannot first describe? Logically,we need to be able to describe exper-tise before attempting to explain it. Thatis why we tried to explicate descriptionbefore attempting to deal with explana-tion. The more specific and detailed thedescription of a phenomenon, the bet-ter we understand it. So, given the rightdescription of expertise, what are we miss-ing? Reductionism asserts that explana-tions go beyond mere description to findmore fundamental causes of the targetbehavior. The causes lie in more basic sci-ences. For psychological phenomena theimmediate causes are likely to be biologi-cal. That’s why the brain is often invokedas the controller, monitor, or generator ofbehavior.

Does brain activity then explain behav-ior? Does the explanation of expertise liein brain circuitry or in genes? The correla-tions are there, between brain activity andbehavior. But saying my brain made me doit is akin to saying “The Devil made medo it.” It’s attributing a cause where thereis no causal evidence. The available scien-tific evidence is strictly correlational. Noone has yet demonstrated that the inde-pendent creation of a brain process willresult in the specific behavior for which itis claimed to be the cause. Thus, assertingthat the brain causes behavior is a matterof faith or belief. And faith has no place inscience. Remember that correlation doesnot imply causation. Neither does correla-tion imply explanation. There is no goodreason other than faith to believe that theexplanation of behavior lies in biologicalevents. The claim that psychological pro-cesses or behaviors cause biological eventsto be what they are is just as plausible orbelievable.

So, in our view, a scientific explana-tion (or “deep understanding”) of exper-tise, based on other sciences, remains tobe realized. We suggest that, if thoroughand complete descriptions of specific casesof expertise can be achieved, then theremight be nothing left to explain, at leastnot in these cases. This possibility suggeststhat, among other things, the implied dif-ference among the three goals of science(definition, description, explanation) is anillusion. Proper and complete descriptionmight supersede the need to explain.

Frontiers in Psychology | Cognition March 2014 | Volume 5 | Article 186 | 212




Bourne et al. Expertise: defined, described, explained

But, if an explanation is to be sought,it will be found, in our view, in thedomain of psychology, rather than somephysical or biological science. By this psy-chological explanation, expertise resultsfrom practice and experience, built ona foundation of talent, or innate ability.The psychology laboratory has revealedempirically based training principles thatfurther elucidate the explanation of exper-tise. These principles enable learnersto maximize the acquisition, retention,and transfer of knowledge and skills,as summarized in Healy and Bourne(2012).

AUTHOR CONTRIBUTIONSLyle E. Bourne Jr. wrote the initial andfinal drafts of the manuscript. James A.Kole and Alice F. Healy edited the initial

draft of the manuscript and suggestedrevisions of it.

ACKNOWLEDGMENTSWe are indebted to the members of theCenter for Research on Training at theUniversity of Colorado for their help-ful suggestions about this research. Theresearch reported here was supported pri-marily by Army Research Office GrantW911NF-05-1-0153.

REFERENCESBourne, L. E. Jr., and Healy, A. F. (2014). Train

Your Mind for Peak Performance: a Science-Based Approach for Achieving Your Goals.Washington, DC: American PsychologicalAssociation.

Ericsson, K. A., Krampe, R. T., and Tesch-Römer,C. (1993). The role of deliberate practice inthe acquisition of expert performance. Psychol.

Rev. 100, 363–406. doi: 10.1037/0033-295X.100.3.363

Healy, A. F., and Bourne, L. E. Jr. (eds.). (2012).Training Cognition: Optimizing Efficiency,Durability, and Generalizability. New York,NY: Psychology Press.

Received: 09 January 2014; accepted: 15 February 2014;published online: 04 March 2014.Citation: Bourne LE Jr., Kole JA and Healy AF (2014)Expertise: defined, described, explained. Front. Psychol.5:186. doi: 10.3389/fpsyg.2014.00186This article was submitted to Cognition, a section of thejournal Frontiers in Psychology.Copyright © 2014 Bourne, Kole and Healy. This isan open-access article distributed under the terms ofthe Creative Commons Attribution License (CC BY).The use, distribution or reproduction in other forumsis permitted, provided the original author(s) or licen-sor are credited and that the original publicationin this journal is cited, in accordance with acceptedacademic practice. No use, distribution or reproduc-tion is permitted which does not comply with theseterms.

www.frontiersin.org March 2014 | Volume 5 | Article 186 | 213











PERSPECTIVE ARTICLEpublished: 06 August 2014

doi: 10.3389/fpsyg.2014.00857

Studying real-world perceptual expertiseJianhong Shen1, Michael L. Mack 2 and Thomas J. Palmeri 1*

1 Vanderbilt Vision Research Center, Department of Psychology, Vanderbilt University, Nashville, TN, USA2 Center for Learning and Department of Psychology, The University of Texas at Austin, Austin, TX, USA

Edited by:

Michael H. Connors, MacquarieUniversity, Australia

Reviewed by:

Andy Wills, Plymouth University, UKMichael H. Connors, MacquarieUniversity, Australia

*Correspondence:

Thomas J. Palmeri, Vanderbilt VisionResearch Center, Department ofPsychology, Vanderbilt University, 301Wilson Hall, 111 21st Avenue South,Nashville, TN 37203, USAe-mail: [email protected]

Significant insights into visual cognition have come from studying real-world perceptualexpertise. Many have previously reviewed empirical findings and theoretical developmentsfrom this work. Here we instead provide a brief perspective on approaches, considerations,and challenges to studying real-world perceptual expertise.We discuss factors like choosingto use real-world versus artificial object domains of expertise, selecting a target domainof real-world perceptual expertise, recruiting experts, evaluating their level of expertise,and experimentally testing experts in the lab and online. Throughout our perspective, wehighlight expert birding (also called birdwatching) as an example, as it has been used as atarget domain for over two decades in the perceptual expertise literature.

Keywords: perceptual expertise, expertise, learning, categorization, recognition, birding

INTRODUCTIONIn nearly every aspect of human endeavor, we find people whostand out for their high levels of skill and knowledge. We callthem experts. Expertise has been studied in domains rangingfrom chess (Chase and Simon, 1973; Gobet and Charness, 2006;Connors and Campitelli, 2014; Leone et al., 2014) to physics(Chi et al., 1981) to sports (Baker et al., 2003). Perceptualexperts, such as ornithologist, radiologists, and mycologists,are noted for their remarkable ability to rapidly and accu-rately recognize, categorize, and identify objects within somedomain. Understanding the development of perceptual exper-tise is more than characterizing the behavior of individuals withuncanny abilities. Rather, if perceptual expertise is the end-point of the trajectory of normal visual learning, then studyingperceptual experts can provide insights into the general princi-ples, limits, and possibilities of human learning and plasticity(e.g., Gauthier et al., 2010).

Several reviews have highlighted empirical findings andtheoretical developments from research on perceptual exper-tise in various modalities (for visual expertise, see, e.g.,McCandliss et al., 2003; Palmeri and Gauthier, 2004; Palmeri andCottrell, 2009; Richler et al., 2011; for auditory expertise, see,e.g., Chartrand et al., 2008; Holt and Lotto, 2008; for tactileexpertise, see, e.g., Behrmann and Ewell, 2003; Reuter et al.,2012). Here, we instead highlight more practical considera-tions that come with studying perceptual expertise; we highlightvisual expertise because this modality has been most exten-sively studied. We specifically consider some choices that faceresearchers: whether to use real-world or artificial objects, whatdomain of perceptual expertise to study, how to recruit par-ticipants, how to evaluate their expertise, and whether to testin the lab or via the web. Throughout our perspective, weuse birding as an example domain because it has been com-monly used in the literature (e.g., Tanaka and Taylor, 1991;Gauthier et al., 2000; Tanaka et al., 2005; Mack et al., 2007;Mack and Palmeri, 2011).

REAL-WORLD vs. ARTIFICIAL DOMAINS OF EXPERTISEExpertise-related research has been conducted using both artificialand real-world objects. Artificial objects include simple stimulilike line orientations, textures, and colors (e.g., Goldstone, 1998;Mitchell and Hall, 2014), and relatively complex novel stimulilike random dot patterns (Palmeri, 1997), Greebles (Gauthier andTarr, 1997; Gauthier et al., 1998, 1999), and Ziggerins (Wong et al.,2009a). Real-world objects include birds, dogs, cars, and othercategories (Tanaka and Taylor, 1991; Gauthier et al., 2000). Studiesusing artificial objects are often training studies, where researchersrecruit novices and train them to become “experts” in a domain.Changes in behavior or brain activity are measured over the courseof training to understand the development of expertise, makingthese studies longitudinal. The weeks of training used in thesestudies can only be a proxy for the years of experience in real-worlddomains. Because real-world expertise takes so long to develop,most real-world studies are cross-sectional.

An advantage of training studies with artificial objects isthe power to establish causality. Experimenters have precisecontrol over properties of novel objects, relationships betweenthem, and how categories are defined (e.g., Richler and Palmeri,2014). Participants can be randomly assigned to conditions andtraining and testing can be carefully controlled. As one exam-ple, Wong et al. (2009a,b) used novel Ziggerins and trainedpeople in two different ways, one of which mirrored individu-ation required for face recognition, another of which mirroredthe letter recognition demands required for reading. Accord-ingly, the face-like training group showed behavior and brainactivity similar to that seen in face recognition while theletter-like training group showed behavior and brain activitysimilar to that seen in letter recognition. Studies of artifi-cial domains of expertise can provide insights into real-worlddomains.

If researchers are interested in understanding what makesexperts experts, not just investigating limits of experience-related changes, then it is important to complement carefully














Shen et al. Real-world perceptual expertise

controlled laboratory studies using artificial domains with thestudy of real-world experts. Because of their quasi-experimentalnature – recruiting novices and those with varying levels of exper-tise as they occur in the real world – these studies cannot establishunambiguous causal relationships between expertise and behav-ioral or brain changes. Apart from considerations of externalvalidity, studies of real-world experts permit the study of a rangeand extent of expertise that cannot easily be reproduced in thelaboratory. And practically speaking, testing real-world percep-tual experts on real-world perceptual stimuli saves researchers theeffort and expense needed to train participants in an artificialdomain.

Studies using real-world domains also come full circle to informstudies using artificial domains. For example, consider the classicresult of Tanaka and Taylor (1991), reproduced in our own onlinereplication in Figure 1. Bird experts categorized birds (their expertdomain) and dogs (their novice domain). For novices (Rosch et al.,1976), objects are categorized faster at a basic level (dog) than asuperordinate (animal) or subordinate level (blue jay), while forexperts (Tanaka and Taylor, 1991; Johnson and Mervis, 1997),objects are categorized as fast at a subordinate level as a basic level.This entry-level shift (Jolicoeur et al., 1984; see also Tanaka et al.,2005; Mack et al., 2009; Mack and Palmeri, 2011) has been usedas a behavioral marker of expertise in training studies employingartificial domains (Gauthier et al., 2000; Gauthier and Tarr, 2002).

Our group recently reviewed considerations that factor intostudies using artificial domains (Richler and Palmeri, 2014), sohere we focus on real-world domains for the remainder of ourperspective.

DOMAINS OF REAL-WORLD PERCEPTUAL EXPERTISEIn addition to everyday domains of perceptual expertise, like faces(Bukach et al., 2006) and letters (McCandliss et al., 2003), stud-ies have used domains ranging from cars and birds (Gauthieret al., 2000), where expertise is not uncommon, to more spe-cialized and sometimes esoteric domains like latent fingerprintidentification (Busey and Parada, 2010; Dror and Cole, 2010),budgie identification (Campbell and Tanaka, 2014), and chicksexing (Biederman and Shiffrar, 1987). The particular choice ofexpert domain depends on a combination of theoretical goals andpractical considerations.

For example, consider a goal of understanding how the abil-ity to categorize at different levels of abstraction changes withperceptual expertise (Mack and Palmeri, 2011), which impactsunderstanding of how categories are learned, represented, andaccessed. Birding is a useful domain because birders must makesubordinate and sub-subordinate categorizations, sometimes at aglance, and often under less than ideal conditions with poor light-ing and camouflage. Other kinds of bird experts have differentskills: budgie experts (a budgerigar is a bred parakeet) can keenlyidentify unique individuals in cages, but need not have expertisewith other birds, while professional chick sexers can quickly dis-criminate male from female genitalia on chicken hatchlings. Inan entirely different domain, fingerprint experts typically matchlatent prints with a known sample, with both clearly visible, pre-sented side by side, and with time limits imposed by the analyst,not the environment.

FIGURE 1 | Mean correct categorization response times for a novice

domain (dogs) and an expert domain (birds) measured online.

Following Tanaka and Taylor (1991), bird experts were tested in a speededcategory verification task where they categorized images at thesuperordinate (animal ), basic (bird or dog), or subordinate (specific speciesor breed) level. In their novice domain (dogs), a classic basic-leveladvantage was observed, whereby categorization at the basic level wassignificantly faster than the superordinate (t22 = 2.67, p = 0.014) andsubordinate level (t22 = 6.75, p < 0.001). In their expert domain (birds),subordinate categorization was as fast as basic-level categorization(t22 = 0.81, p = 0.429). This replication was conducted using an onlineWordpress + Flash custom website with only 23 participants from a singleshort 10 min experimental session. Error bars represent 95% confidenceintervals on the level × domain interaction.

There are real-world consequences for studying certaindomains of perceptual expertise, such as latent fingerprint exam-ination. Despite the widespread use of forensic evidence – as wellas its popular depiction on television – a recent National ResearchCouncil of the National Academy of Sciences (2009) noted a“dearth of peer-reviewed, published studies establishing the sci-entific bases and validity of many forensic methods,” especiallythose methods that require subjective visual pattern analysis andexpert testimony. That scientific evidence is emerging, especiallyin the case of latent fingerprint expertise (e.g., Busey and Parada,2010; Busey and Dror, 2011).

The choice of domain can also be influenced by various prac-tical considerations. It is easier to study perceptual expertise ina domain with millions of possible participants than an esotericdomain with a few isolated members. It is easier to study a domainwhere relevant stimuli are widely available in books and online.And it is easier to study a domain without barriers to contact,which can be the case for experts in the military, homeland security,and certain professions. For example, studies of expert baggagescreeners require coordination with the Transportation SecurityAdministration (TSA) and many details regarding stimuli andprocedures cannot be shared with the public (e.g., Wolfe et al.,2013). In the case of birding, there are millions of people in theUS alone who consider birding a hobby, spending hours in theiryards and parks, and billions on books, equipment, and travel (LaRouche, 2006). Photos of birds are widely available; books have





been published on particularly difficult bird identifications (e.g.,Kaufman, 1999, 2011). Birders regularly participate in citizen sci-ence efforts, such as the Christmas bird count and provide data onbird sightings to databases like ebird.org. Anecdotally, this trans-lates into a keen interest in science and a willingness to participatein research.

RECRUITINGIn the past, experts usually had to be recruited locally, withadvertisements posted around a university campus and in localnewspapers. It may be hard for some to remember that it has onlybeen in the past several years that not having an email address hasbecome almost equivalent to not having a phone number, and thatonly recently has it become the case that most people have someInternet access. Being able to recruit participants more widely viathe Internet promises not only to increase heterogeneity of par-ticipants, but also, and especially relevant for expertise research,promises to locate participants with a far greater range of exper-tise than might be possible when recruiting in a local geographicregion.

One rapidly exploding means of recruiting and testing (see“Testing”) participants is Amazon Mechanical Turk (AMT). AMTallows hundreds of subjects to be easily recruited and tested in amatter of days; participants on AMT are more demographicallydiverse than typical American college samples (Buhrmester et al.,2011). This diversity is important for research examining individ-ual differences in perception and cognition. While the potentialpopulation of AMT workers is large, it is unknown how manywith high levels of domain expertise might be workers on the plat-form. For expertise research, recruitment via AMT may need to besupplemented by more direct recruitment of true domain experts(e.g., Van Gulick, 2014).

Large domains of expertise have organizations, web sites,blogs, and even tweets and Facebook updates that target par-ticular individuals. In principle, online recruiting through thesechannels offers a quick, easy, and inexpensive means of findingexperts. These could involve paid advertisements online and inelectronic newsletters. More directly, these could involve mes-sages sent to email lists. The biggest challenge to this, however,is that many professional organizations or workplaces wouldrarely allow, and many outright prohibit, direct solicitation ofmembers or employees, even for basic research; researchers can-not directly contact TSA baggage screeners or latent fingerprintexaminers. By comparison, birding organizations, including localOrnithological and Audubon Societies, whose members joinas part of a hobby, not a profession, can be less restrictivein terms of allowing contact with members, so long as con-tact is non-intrusive. In our case, we have identified severalhundred birding groups in the US and Canada, we have con-tacted several dozen directly, and have received permission tosolicit volunteer participants from most, having so far testedseveral hundred birders with a wide range of experience andexpertise.

EVALUATING LEVELS OF PERCEPTUAL EXPERTISEHow do we know someone is a perceptual expert? A simpleapproach relies on subjective self-rating, often supplemented by

self-report on the amount of formal training, years of expe-rience, or community reputation. For example, bird expertsin Tanaka and Taylor (1991) were recommended by mem-bers of bird-watching organizations and had a minimum of10 years of experience, and those in Johnson and Mervis(1997) led birding field trips and some had careers related tobirding.

It is now well-recognized that self-reports of expertise are insuf-ficient and that objective measures of expert performance areneeded (e.g., Ericsson, 2006, 2009); self-report measures of per-ceptual expertise are not always good predictors of performance(e.g., McGugin et al., 2012; Van Gulick, 2014). Therefore, recentwork has used quantitative measures to assess expert abilities (e.g.,see Gauthier et al., 2010). A detailed review and discussion of suchmeasures is well beyond the scope of a brief perspective piece.A variety of quantitative measures of perceptual expertise havebeen used and new measures are currently being developed – theseefforts to develop and validate new measures reflect a quickly grow-ing interest in exploring individual difference in visual cognition(e.g., Wilmer et al., 2010; Gauthier et al., 2013; Van Gulick, 2014).

While expert-novice differences are sometimes looselydescribed as if they were dichotomous, it is self-evident thatexpertise is a continuum, people vary in their level of exper-tise, and any measure of expertise must place individuals alonga (perhaps multidimensional) continuum. Some behavioral orneural markers might distinguish pure novices from those withsome experience but asymptote at only an intermediate levelof expertise, while other behavioral or neural markers mightdistinguish the true experts from more middling experts andnovices. Understanding the continuum of behavioral and brainchanges, whether they are asymptotic, monotonic, or evennon-monotonic over the continuum of expertise, can haveimportant implications for understanding mechanistically andcomputationally how perceptual expertise develops (e.g., seePalmeri et al., 2004).

Briefly, one useful measure has focused on the perceptualpart of perceptual expertise: using a simple one-back match-ing task, images are presented one at a time and participantsmust say whether consecutive pictures are the same or differ-ent. Experts have higher discriminability (d′) on images fromtheir domain of expertise relative to non-expert domains, andthis difference predicts behavioral and brain differences (e.g.,Gauthier et al., 2000; Gauthier and Tarr, 2002). Another mea-sure has focused on memory as an index of perceptual expertise:the Vanderbilt Expertise Task (VET; McGugin et al., 2012) mir-rors aspects of the Cambridge Face Memory task (Duchaine andNakayama, 2006). Participants memorize exemplars from sev-eral different artifact and natural categories and then recognizeother instances under a variety of conditions, and these differ-ences in memory within particular domains predict behavioraland brain differences (e.g., McGugin et al., 2014). With our inter-est in categorization at different levels of abstraction, in work inpreparation, we have developed a measure that has focused oncategorical knowledge in perceptual expertise: adapting commonpsychometric approaches, we are refining what could essentiallybe characterized as an Scholastic Assessment Test (SAT, a stan-dardized test widely used for college admission in the United





States) of birding knowledge, with multiple-choice identificationsof bird images ranging from easy (common backyard birds likethe Blue Jay), to intermediate (distinctive yet far less commonbirds, like the Pileated Woodpecker or Great Kiskadee), to quitedifficult identifications that even fairly expert birders find diffi-cult (like discriminating Bohemian from Cedar Waxwing, Hairyfrom Downy Woodpecker, or correctly identifying the manyextremely similar warblers, sparrows, or flycatchers). Future workmust consider to what extent different measures of perceptualexpertise capture the same dimensions of expert knowledge andpredict the same behavioral and brain measures that vary withexpertise.

TESTINGLaboratory testing allows careful control and monitoring ofperformance, permits experiments that require precisely-timedstimulus presentations, and of course allows sophisticated behav-ioral and brain measures like eye movements, fMRI, EEG, andthe like. But laboratory testing incurs a potential cost in thatthe number of laboratory participants is often limited due to theexpense of subject reimbursement, personnel hours, lab space,and equipment. And for any study of unique populations whomight be geographically dispersed, such as perceptual experts, thecost of bringing participants to the laboratory can be prohibitivelyexpensive.

Until fairly recently, the only real methods for testing partic-ipants from a wide geographic area, apart from having experi-menters or participants travel, was to have the experiments travel.For simple studies, this could mean mailed pencil-and-paper tests,while for more sophisticated studies, this could mean sendingdisks or CDs to participants to run on a home computer (e.g.,Tanaka et al., 2010). As anyone who programs well knows, gettingsoftware to run properly on a wide range of computer hardwareand operating system versions can be a daunting task. In the pastfew years, it has become popular, and wildly successful, to haveexperiments run via a web browser. While not entirely immune tothe vagaries of hardware and operating system versions, browser-based applications are often more robust to significant variation,and can often automatically prompt users for upgrades to requisitesoftware plug-ins.

There are multiple platforms and approaches to online web-based experiments. One approach, highlighted earlier, uses AMT.In AMT, researchers publish Human Intelligence Tasks (HITs) thatregistered workers can complete in exchange for modest monetarycompensation. AMT integrates low-level programming tools forstimulus creation, test design, and programming into one web-based application; other elements in AMT include automatedcompensation, recruitment, and data collection. Aside from theavailability of these tools, a clear advantage of AMT is the poten-tial to recruit from a large and diverse pool of participants. Analternative approach is to develop and support a custom web-based server for experiments. There are powerful tools for creatingweb pages, such as Wordpress (wordpress.org), and fairly sophis-ticated programs can be developed in Adobe Flash or Javascript(e.g., De Leeuw, 2014; Simcox and Fiez, 2014). Perhaps an advan-tage of such custom portals is that people may be more attractedto them because of their interest in participating in research, not

because of the potential to earn money, as might sometimes bethe case for AMT. In the end, we suspect that most labs willuse a combination of both platforms for recruiting, testing, orboth.

At least given current computer hardware in wide use, apotential vexing problem for web-based experiments is tim-ing. Fortunately, platforms such as Flash and Javascript runon the local (participant) computer, so properly-designed pro-grams can avoid problems that could be introduced by variabilityin Internet connection speeds. Thankfully, reasonable responsetime measurements can be obtained (Reimers and Stewart, 2007;Crump et al., 2013; Simcox and Fiez, 2014). Indeed, as illus-trated in Figure 1, we have successfully observed differencesin RTs for expert and novice domains in online experimentsusing a Wordpress + Flash environment that mirror observationsof expert speeded categorization from classic laboratory studies(Tanaka and Taylor, 1991). Unfortunately, the most critical lim-itation for now concerns stimulus timing. It is well known thatLCD monitors in wide use have response characteristics far toosluggish to permit the kind of “single-refresh” presentations thatwould have been possible on previous CRTs. While presentationtimes of 100 ms or more are probably a safe bet, anything fasterwould require calibration to check that a participant had a suffi-ciently responsive monitor; it may be that the next generation ofLCD, LED, or other technologies will (hopefully) eliminate theselimitations.

SUMMARYMost human endeavors have a perceptual component. For exam-ple, keen visual perception is required in sports, medicine, science,games like chess, and a wide range of skilled behavior. Thusresearch on real-world perceptual expertise has potential theo-retical and applied impacts to many domains. Here we brieflyoutlined at least some of the practical considerations that factorinto research on real-world perceptual expertise. Several of theseconsiderations are things that researchers often fret over behindthe scenes without making it into a typical research publication,so in that sense we hope this brief perspective fills a small butimportant hole in the literature.

ACKNOWLEDGMENTSThis research was supported in part by grants SBE-1257098and SMA-1041755 from the National Science Foundation and aDiscovery Grant from Vanderbilt University.

REFERENCESBaker, J., Cote, J., and Abernethy, B. (2003). Sport-specific practice and the devel-

opment of expert decision-making in team ball sports. J. Appl. Sport Psychol. 15,12–25. doi: 10.1080/10413200305400

Behrmann, M., and Ewell, C. (2003). Expertise in tactile pattern recognition. Psychol.Sci. 14, 480–492. doi: 10.1111/1467-9280.02458

Biederman, I., and Shiffrar, M. M. (1987). Sexing day-old chicks: a case study andexpert systems analysis of a difficult perceptual-learning task. J. Exp. Psychol.Learn. Mem. Cogn. 13, 640–645. doi: 10.1037//0278-7393.13.4.640

Buhrmester, M., Kwang, T., and Gosling, S. D. (2011). Amazon’s Mechanical Turk:a new source of inexpensive, yet high-quality, data? Perspect. Psychol. Sci. 6, 3–5.doi: 10.1177/1745691610393980

Bukach, C. M., Gauthier, I., and Tarr, M. J. (2006). Beyond faces and modular-ity: the power of an expertise framework. Trends Cogn. Sci. 10, 159–166. doi:10.1016/j.tics.2006.02.004





Busey, T. A., and Dror, I. E. (2011). “Special abilities and vulnerabilities in forensicexpertise,” in Friction Ridge Sourcebook, ed. A. McRoberts (Washington, DC: NIJPress), 1–23.

Busey, T. A., and Parada, F. J. (2010). The nature of expertise in fingerprintexaminers. Psychon. Bull. Rev. 17, 155–160. doi: 10.3758/PBR.17.2.155

Campbell, A., and Tanaka, J. (2014). “Testing the face-specificity of the inversioneffect in budgie experts,” in Poster session presented at the Fourteenth AnnualMeeting of the Vision Sciences Society, St. Pete Beach, FL.

Chartrand, J. P., Peretz, I., and Belin, P. (2008). Auditory recognition expertise anddomain specificity. Brain Res. 1220, 191–198. doi: 10.1016/j.brainres.2008.01.014

Chase, W. G., and Simon, H. A. (1973). Perception in chess. Cogn. Psychol. 4, 55–81.doi: 10.1016/0010-0285(73)90004-2

Chi, M. T., Feltovich, P. J., and Glaser, R. (1981). Categorization and represen-tation of physics problems by experts and novices. Cogn. Sci. 5, 121–152. doi:10.1207/s15516709cog0502_2

Connors, M. H., and Campitelli, G. (2014). Expertise and the representation ofspace. Front. Psychol. 5:270. doi: 10.3389/fpsyg.2014.00270

Crump, M. J., McDonnell, J. V., and Gureckis, T. M. (2013). Evaluating Ama-zon’s Mechanical Turk as a tool for experimental behavioral research. PLoS ONE8:e57410. doi: 10.1371/journal.pone.0057410

De Leeuw, J. R. (2014). jsPsych: a JavaScript library for creating behavioral exper-iments in a web browser. Behav. Res. Methods doi: 10.3758/s13428-014-0458-y[Epub ahead of print].

Dror, I. E., and Cole, S. A. (2010). The vision in “blind” justice: expert perception,judgment, and visual cognition in forensic pattern recognition. Psychon. Bull.Rev. 17, 161–167. doi: 10.3758/PBR.17.2.161

Duchaine, B., and Nakayama, K. (2006). The Cambridge Face Memory Test: resultsfor neurologically intact individuals and an investigation of its validity usinginverted face stimuli and prosopagnosic participants. Neuropsychologia 44, 576–585. doi: 10.1016/j.neuropsychologia.2005.07.001

Ericsson, K. A. (2006). “The influence of experience and deliberate practice on thedevelopment of superior expert performance,” in The Cambridge Handbook ofExpertise and Expert Performance, eds K. A. Ericsson, N. Charness, P. Feltovich,and R. R. Hoffman (Cambridge, UK: Cambridge University Press), 691–698. doi:10.1017/CBO9780511816796.038

Ericsson, K. A. (2009). “Enhancing the development of professional perfor-mance: implications from the study of deliberate practice,” in The Devel-opment of Professional Expertise: Toward Measurement of Expert Performanceand Design of Optimal Learning Environments, ed. K. A. Ericsson (NewYork, NY: Cambridge University Press), 412–425. doi: 10.1017/CBO9780511609817.022

Gauthier, I., McGugin, R., Richler, J. J., Herzmann, G., Speegle, M., and Van Gulick,A. E. (2013). Experience with objects moderates the overlap between object andface recognition performance, suggesting a common ability. J. Vis. 13, 982–982.doi: 10.1167/13.9.982

Gauthier, I., Skudlarski, P., Gore, J. C., and Anderson, A. W. (2000). Expertise forcars and birds recruits brain areas involved in face recognition. Nat. Neurosci. 3,191–197. doi: 10.1038/72140

Gauthier, I., and Tarr, M. J. (1997). Becoming a “Greeble” expert: exploringmechanisms for face recognition. Vision Res. 37, 1673–1682. doi: 10.1016/S0042-6989(96)00286-6

Gauthier, I., and Tarr, M. J. (2002). Unraveling mechanisms for expert objectrecognition: bridging brain activity and behavior. J. Exp. Psychol. Hum. Percept.Perform. 28, 431–446. doi: 10.1037//0096-1523.28.2.431

Gauthier, I., Tarr, M. J., Anderson, A. W., Skudlarski, P., and Gore, J. C.(1999). Activation of the middle fusiform “face area” increases with exper-tise in recognizing novel objects. Nat. Neurosci. 2, 568–573. doi: 10.1038/9224

Gauthier, I., Tarr, M. J., and Bub, D. N. (eds). (2010). Perceptual Expertise:Bridging Brain and Behavior. New York, NY: Oxford University Press. doi:10.1093/acprof:oso/9780195309607.001.0001

Gauthier, I., Williams, P., Tarr, M. J., and Tanaka, J. (1998). Training “greeble”experts: a framework for studying expert object recognition processes. Vision Res.38, 2401–2428. doi: 10.1016/S0042-6989(97)00442-2

Gobet, F., and Charness, N. (2006). “Expertise in chess,” in The Cambridge Handbookof Expertise and Expert Performance, eds K. A. Ericsson, N. Charness, P. Feltovich,and R. R. Hoffman (Cambridge, UK: Cambridge University Press), 523–538. doi:10.1017/CBO9780511816796.038

Goldstone, R. L. (1998). Perceptual learning. Annu. Rev. Psychol. 49, 585–612. doi:10.1146/annurev.psych.49.1.585

Holt, L. L., and Lotto, A. J. (2008). Speech perception within an auditory cognitivescience framework. Curr. Dir. Psychol. Sci. 17, 42–46. doi: 10.1111/j.1467-8721.2008.00545.x

Johnson, K. E., and Mervis, C. B. (1997). Effects of varying levels of expertiseon the basic level of categorization. J. Exp. Psychol. Gen. 126, 248–277. doi:10.1037/0096-3445.126.3.248

Jolicoeur, P., Gluck, M. A., and Kosslyn, S. M. (1984). Pictures and names: mak-ing the connection. Cogn. Psychol. 16, 243–275. doi: 10.1016/0010-0285(84)90009-4

Kaufman, K. (1999). A Peterson Field Guide to Advanced Birding: Birding Challengesand How to Approach Them. Boston, MA: Houghton Mifflin Harcourt.

Kaufman, K. (2011). Kaufman Field Guide to Advanced Birding: Understanding WhatYou See and Hear. New York, NY: Houghton Mifflin Harcourt.

La Rouche, G. P. (2006). “Birding in the united states: a demographic and economicanalysis,” in Waterbirds Around the World, eds G. C. Boere, C. A. Galbraith, andD. A. Stroud (Edinburgh, UK: The Stationery Office), 841–846.

Leone, M. J., Fernandez Slezak, D., Cecchi, G. A., and Sigman, M. (2014).The geometry of expertise. Front. Psychol. 5:47. doi: 10.3389/fpsyg.2014.00047

Mack, M. L., and Palmeri, T. J. (2011). The timing of visual object categorization.Front. Psychol. 2:165. doi: 10.3389/fpsyg.2011.00165

Mack, M. L., Wong, A. C.-N., Gauthier, I., Tanaka, J. W., and Palmeri, T. J. (2007).“Unraveling the time-course of perceptual categorization: does fastest mean first?”in Proceedings of the 29th Annual Conference of the Cognitive Science Society,Mahwah, NJ, 1253–1258.

Mack, M. L., Wong, A. C.-N., Gauthier, I., Tanaka, J. W., and Palmeri, T.J. (2009). Time course of visual object categorization: fastest does not nec-essarily mean first. Vision Res. 49, 1961–1968. doi: 10.1016/j.visres.2009.05.005

McCandliss, B. D., Cohen, L., and Dehaene, S. (2003). The visual word form area:expertise for reading in the fusiform gyrus. Trends Cogn. Sci. 7, 293–299. doi:10.1016/S1364-6613(03)00134-7

McGugin, R. W., Richler, J. J., Herzmann, G., Speegle, M., and Gauthier,I. (2012). The vanderbilt expertise test reveals domain-general and domain-specific sex effects in object recognition. Vision Res. 69, 10–22. doi:10.1016/j.visres.2012.07.014

McGugin, R. W., Van Gulick, A. E., Tamber-Rosenau, B. J., Ross, D. A., and Gau-thier, I. (2014). Expertise effects in face-selective areas are robust to clutter anddiverted attention, but not to competition. Cereb. Cortex doi: 10.1093/cercor/bhu060 [Epub ahead of print].

Mitchell, C., and Hall, G. (2014). Can theories of animal discrimination explainperceptual learning in humans? Psychol. Bull. 140, 283–307. doi: 10.1037/a0032765

National Research Council of the National Academy of Sciences. (2009). Strength-ening Forensic Science in the United States: A Path Forward. Washington, DC:National Academies Press.

Palmeri, T. J. (1997). Exemplar similarity and the development of automaticity.J. Exp. Psychol. Learn. Mem. Cogn. 23, 324–54. doi: 10.1037//0278-7393.23.2.324

Palmeri, T. J., and Cottrell, G. W. (2009). “Modeling Perceptual Expertise,”in Perceptual Expertise: Bridging Brain and Behavior, eds D. N. Bub, M. J.Tarr, and I. Gauthier (New York, NY: Oxford University Press), 197–245. doi:10.1093/acprof:oso/9780195309607.003.0008

Palmeri, T. J., and Gauthier, I. (2004). Visual object understanding. Nat. Rev.Neurosci. 5, 291–303. doi: 10.1038/nrn1364

Palmeri, T. J., Wong, A. C.-N., and Gauthier, I. (2004). Computational approachesto the development of perceptual expertise. Trends Cogn. Sci. 8, 378–386. doi:10.1016/j.tics.2004.06.001

Reimers, S., and Stewart, N. (2007). Adobe Flash as a medium for online experi-mentation: A test of reaction time measurement capabilities. Behav. Res. Methods39, 365–370. doi: 10.3758/BF03193004

Reuter, E. M., Voelcker-Rehage, C., Vieluf, S., and Godde, B. (2012). Touch percep-tion throughout working life: Effects of age and expertise. Exp. Brain Res. 216,287–297. doi: 10.1007/s00221-011-2931-5

Richler, J. J., and Palmeri, T. J. (2014). Visual category learning. WIREs Cogn. Sci. 5,75–94. doi: 10.1002/wcs.1268





Richler, J. J., Wong, Y. K., and Gauthier, I. (2011). Perceptual expertise as a shift fromstrategic interference to automatic holistic processing. Curr. Dir. Psychol. Sci. 20,129–134. doi: 10.1177/0963721411402472

Rosch, E., Mervis, C. B., Gray, W. D., Johnson, D. M., and Boyes-Braem, P. (1976).Basic objects in natural categories. Cogn. Psychol. 8, 382–439. doi: 10.1016/0010-0285(76)90013-X

Simcox, T., and Fiez, J. A. (2014). Collecting response times using AmazonMechanical Turk and Adobe Flash. Behav. Res. Methods 46, 95–111. doi:10.3758/s13428-013-0345-y

Tanaka, J. W., Curran, T., and Sheinberg, D. L. (2005). The training and transferof real-world perceptual expertise. Psychol. Sci. 16, 145–151. doi: 10.1111/j.0956-7976.2005.00795.x

Tanaka, J. W., and Taylor, M. (1991). Object categories and expertise: Is the basiclevel in the eye of the beholder? Cogn. Psychol. 23, 457–482. doi: 10.1016/0010-0285(91)90016-H

Tanaka, J. W., Wolf, J. M., Klaiman, C., Koenig, K., Cockburn, J., Herlihy, L., et al.(2010). Using computerized games to teach face recognition skills to children withautism spectrum disorder: the Let’s Face It! program. J. Child Psychol. Psychiatry51, 944–952. doi: 10.1111/j.1469-7610.2010.02258.x

Van Gulick, A. E. (2014). Measurement of Semantic Knowledge and its Contributionto Object Recognition Performance. Doctoral Dissertation, Vanderbilt University,Nashville, TN.

Wilmer, J. B., Germine, L., Chabris, C. F., Chatterjee, G., Williams, M., Loken, E.,et al. (2010). Human face recognition ability is specific and highly heritable. Proc.Natl. Acad. Sci. U.S.A. 107, 5238–5241. doi: 10.1073/pnas.0913053107

Wolfe, J. M., Brunelli, D. N., Rubinstein, J., and Horowitz, T. S. (2013). Prevalenceeffects in newly trained airport checkpoint screeners: trained observers miss raretargets, too. J. Vis. 13, 1–9. doi: 10.1167/13.3.33

Wong, A. C.-N., Palmeri, T. J., and Gauthier, I. (2009a). Conditions for facelikeexpertise with objects: becoming a ziggerin expert—but which type? Psychol. Sci.20, 1108–1117. doi: 10.1111/j.1467-9280.2009.02430.x

Wong, A. C.-N., Palmeri, T. J., Rogers, B. P., Gore, J. C., and Gauthier, I.(2009b). Beyond shape: how you learn about objects affects how they arerepresented in visual cortex. PLoS ONE 4:e8405. doi: 10.1371/journal.pone.0008405


Received: 31 May 2014; accepted: 19 July 2014; published online: 06 August 2014.Citation: Shen J, Mack ML and Palmeri TJ (2014) Studying real-world perceptualexpertise. Front. Psychol. 5:857. doi: 10.3389/fpsyg.2014.00857This article was submitted to Cognition, a section of the journal Frontiers in Psychology.Copyright © 2014 Shen, Mack and Palmeri. This is an open-access article distributedunder the terms of the Creative Commons Attribution License (CC BY). The use, dis-tribution or reproduction in other forums is permitted, provided the original author(s)or licensor are credited and that the original publication in this journal is cited, inaccordance with accepted academic practice. No use, distribution or reproduction ispermitted which does not comply with these terms.






REVIEW ARTICLEpublished: 16 October 2014

doi: 10.3389/fpsyg.2014.01155

Metacognition and action: a new pathway to understandingsocial and cognitive aspects of expertise in sportTadhg E. MacIntyre1*, Eric R. Igou 2 , Mark J. Campbell 1, Aidan P. Moran 3 and James Matthews 4

1 Department of Physical Education and Sport Sciences, University of Limerick, Limerick, Ireland2 Department of Psychology, University of Limerick, Limerick, Ireland3 School of Psychology, University College Dublin, Dublin, Ireland4 School of Public Health, Physiotherapy and Population Science, University College Dublin, Dublin, Ireland

Edited by:


Reviewed by:

Derrick L. Hassert, Trinity ChristianCollege, USATom Carr, Michigan State University,USAGuillermo Macbeth, NationalScientific and Technical ResearchCouncil, Argentina

*Correspondence:

Tadhg E. MacIntyre, Department ofPhysical Education and SportSciences, University of Limerick,PESS Building, County Limerick,Limerick, Irelande-mail: [email protected]

For over a century, psychologists have investigated the mental processes of expert perform-ers – people who display exceptional knowledge and/or skills in specific fields of humanachievement. Since the 1960s, expertise researchers have made considerable progressin understanding the cognitive and neural mechanisms that underlie such exceptionalperformance. Whereas the first modern studies of expertise were conducted in relativelyformal knowledge domains such as chess, more recent investigations have exploredelite performance in dynamic perceptual-motor activities such as sport. Unfortunately,although these studies have led to the identification of certain domain-free generalizationsabout expert-novice differences, they shed little light on an important issue: namely,experts’ metacognitive activities or their insights into, and regulation of, their own mentalprocesses. In an effort to rectify this oversight, the present paper argues that metacognitiveprocesses and inferences play an important if neglected role in expertise. In particular,we suggest that metacognition (including such processes as “meta-attention,” “meta-imagery” and “meta-memory,” as well as social aspects of this construct) provides awindow on the genesis of expert performance. Following a critique of the standardempirical approach to expertise, we explore some research on “metacognition” and“metacognitive inference” among experts in sport. After that, we provide a brief evaluationof the relationship between psychological skills training and metacognition and commenton the measurement of metacognitive processes. Finally, we summarize our conclusionsand outline some potentially new directions for research on metacognition in action.

Keywords: metacognition, expertise, cognition, motor cognition, social cognition, cognitive neuroscience, sport,

sport psychology

INTRODUCTIONExpertise is characterized by superior reproducible performancesand “refers to the characteristics, skills, and knowledge that distin-guish experts from novices and less experienced people” (Ericsson,2006a, p. 3). Quintessentially, sport provides many such instances.For example, when Lewis (Lewis and Marx, 1996) became the firsttrack and field athlete to win four consecutive Olympic titles, heaccomplished this feat with a long jump of 8 m 50 cm, winningby a margin of 21 cm. Sport is a domain that provides bench-marks to distinguish experts from novices, through performanceoutcomes (e.g., podium placing), player statistics (e.g., battingaverage in baseball) or competition level (e.g., Olympic vs. Colle-giate). Given such criteria, it is not surprising that the question ofhow one becomes an expert within the sport domain has been ofincreasing scientific (Ericsson, 2014; Hambrick et al., 2014a) andpopular interest (Ross, 2006; Gladwell, 2008) in the two decadessince Ericsson’s seminal paper on deliberate practice. “Deliberatepractice presents performers with tasks that are initially outsidetheir current realm of reliable performance, yet can be masteredwithin hours of practice by concentrating on critical aspects andby gradually refining performance through repetitions after feed-back” (Ericsson, 2006b, p. 692). Although subject of much debate

(e.g., Baker and Young, 2014; Detterman, 2014), the theory ofdeliberate practice and the development of expertise both war-rant further analyses. This paper adds to the expertise debateby presenting a novel argument contending that metacognitiveprocesses are central to expertise in the sport context. Further-more, we suggest that some of the aforementioned controversiesin the research literature, that stem from conflict between thosefocused largely on exploring the role of automaticity in skilledperformance, and researchers focused on understanding the rep-resentation of experts knowledge, may have led to the explanatoryrole of metacognition being overlooked.

In this review we propose that our understanding of aspectsof both social and cognitive dimensions of sporting expertisecan be adequately explained from a metacognitive perspective.The potential of metacognitive inferences and domain-generalskills including psychological skills training (PST) are positedas integral to the genesis of expert performance. Subsequently,the contribution of both mental imagery (e.g., mental practice)and attentional strategies (e.g., routines) to our understandingof expertise and metacognition will be discussed. Finally, newdirections for future research that emanate from our metacog-nitive perspective on sporting expertise will be outlined. Firstly,

www.frontiersin.org October 2014 | Volume 5 | Article 1155 | 220















MacIntyre et al. Metacognition and sporting expertise

however, in the rationale for our approach we will attempt toanswer the following questions: are there limits to the currentexpertise approach? Why is sport an appropriate field of study?And finally, what is metacognition?

ARE THERE LIMITS TO THE EXPERTISE APPROACH?Historically, the rich tapestry of research on expert performancehas been interwoven with a common thread-the study of grand-masters in chess (Williams and Ericsson, 2005). Investigationsinto expertise in chess, a competitive sporting activity that wasrule bound, amenable to measurement through objective rat-ings (e.g., ELO rankings), with a range of possible contextualrequirements (e.g., blindfold chess; Campitelli and Gobet, 2005)led to a proliferation of literature on the topic (de Groot, 1965;Chase and Simon, 1973; Holding, 1985, 1992). Chess is a chal-lenge of perceptual-cognitive skill and thus provides a fittinglaboratory for testing constructs such as pattern recognition,visual imagery, and memory (Bilalic et al., 2010). Sport in themore traditional sense emphasizes motor skill execution understressful conditions typically in a dynamic environment (Bakerand Young, 2014). One legacy of the chess expertise literaturewas that this perceptual-cognitive lens was subsequently appliedin the sport domain (MacIntyre et al., 2013). Two interlinkedevents led researchers to become enthusiastic in their study ofvisual cognition and sport (Williams and Ford, 2007). Firstly, theemergence of the expert performance approach (Ericsson andSmith, 1991) and later, the theory of deliberate practice (Eric-sson et al., 1993), both provided an impetus for investigationsinto perceptual-cognitive skills, such as anticipation and decision-making in sport. The tenet that sport could be a dynamic naturallaboratory was well made (Moran, 1996; Williams and Erics-son, 2005) and the development of innovative methodologiesoccurred in parallel (e.g., eye-tracking as a measure of atten-tion). A burgeoning literature developed and sport as a domain ofstudy gained popularity as a result (Moran, 2009; MacIntyre et al.,2013).

However, there were limits to this approach, particularly inthe focus on visual-cognitive expertise, which arguably was to thedetriment of our understanding of the underlying psychologicalprocesses. Take, for example, the quiet eye phenomenon which hasrecently gained prominence in sport science research (Vine et al.,2014). Increasingly, this is becoming a topic of interest within bothcognitive psychology (Klostermann et al., 2013) and neuroscience(Vine et al., 2011). The quiet eye is defined as “a final fixation ortracking gaze that is located on a specific location or object inthe visuomotor workspace within 3◦ of visual angle (or less) for aminimum of 100 ms” (Vickers, 2007, p. 11) prior to the onset ofa critical movement. According to Vickers, quiet eye offset occurswhen the gaze deviates off a specific location for more than 100 ms(Vickers, 2007, p. 11). Despite the success in establishing a quieteye phenomenon “there has been a lack of explicit tests of the pro-cesses through which quiet eye training interventions exert theirpositive effect” (Vine et al., 2014, p. S237). To date, little knowl-edge of the psychological basis of the quiet eye phenomenon hasemerged (Moran, 2012a). Similarly, while a wealth of knowledgehas accumulated on the characteristics of individuals’ saccadicpursuit during visual attention tasks (Williams and Ford, 2007),

little evidence exists to support the trainability of visual search(Mann et al., 2007).

A further limitation to the expertise approach is that, for exam-ple, the focus has been on a narrow set of conclusions fromthe original publication on deliberate practice (Ericsson et al.,1993). Therefore, it is not surprising that the debate over thecontribution of deliberate practice to expert performance con-tinues in chess (Campitelli and Gobet, 2011; Detterman, 2014),sporting expertise (Williams and Ford, 2007; Baker and Young,2014) and professional expertise (Ericsson, 2009; Hoffman, 2014).Disagreements over the number of hours accumulated, start-ing age, and the link to general cognitive abilities continue todominate the field (e.g., Gobet and Campitelli, 2007; Baker andYoung, 2014; Hambrick et al., 2014a). For example, Hambricket al. (2014b, p. 113) concluded that evidence had accumulatedto suggest that, although relevant, deliberate practice, in itself,“does not largely account for individual differences in perfor-mance.” One caveat to be considered here is that these authorswere concerned with the predictive ability of the theory solelywithin the context of chess. The acquisition of motor skills inthe traditional sporting context is arguably more complex (Vosset al., 2010). For example, even defining deliberate practice amongathletes is more challenging than with chess grandmasters (Mac-Intyre et al., 2013; Healy et al., 2014). Nevertheless, the conclusionby Hambrick et al. (2014b, p. 114) that “the question now iswhat else matters” suggests that we should consider a broaderrange of constructs in order to more comprehensively understandexpertise.

Before we consider the construct of metacognition, the ratio-nale for studying athletes and sport performers must be madereadily apparent. Numerous authors have highlighted the rolein which sport can provide a natural laboratory for the study ofconstructs within psychology and expertise (Ericsson and Smith,1991; Moran, 1996; MacIntyre et al., 2013). According to Erics-son (2009, p. 18) “performance can be publically observed andeven objectively measured in open competition and public perfor-mances.” Similarly, it has been noted that the high performancesport environment is dynamic. For example, typically perform-ers have to execute complex skills under conditions of extremestress where their limits are being constantly challenged (Bakerand Young, 2014). Among the topics that have only recentlyreceived scrutiny are the role of attention and the allocationof effort in deliberate practice (Baker and Young, 2014). Oneexplanation for this is that researchers concentrated on the vari-ables that were most measurable, including the quantification ofhours in practice (Helsen et al., 1998). A challenge for researchershas been reconciling the automaticity and procedural knowl-edge, central to expert sport performance, with the notion thatdeclarative knowledge and metacognitive abilities may also play arole in the acquisition of expertise (Stanley and Krakauer, 2013;Toner, 2014). To explain, while procedural knowledge is inherentlylinked to optimum sport performance, declarative knowledge mayhave both a debilitative (Beilock and Carr, 2001) and facilitativerole (Carson and Collins, 2011; MacIntyre et al., 2013; Brick et al.,2014). For example, it is probable that Carl Lewis knew his precisestride count to enable him to hit the board and take off into thesandpit at the 1984 Olympic Games. Thus, expertise in sports goes

Frontiers in Psychology | Cognition October 2014 | Volume 5 | Article 1155 | 221




beyond mere procedural knowledge and arguably metacognitiveprocesses are present at all stages of the target skill and may workin parallel. We thus propose an integrative model of expertise insports, one that explores action and cognition in sport, a topic thathas arguably only recently returned to the forefront of psychology.

In Rosenbaum (2005) suggested that researchers in psychologyhave historically turned relatively late to cognition and action.While this point is debatable, given the recent emergence ofresearch on exceptional performance states (e.g., Choking; Beilockand Carr, 2001), the paucity of research on action in priordecades may be worthy of review. Understanding movement hadlong been the preserve of the fields of motor control, biome-chanics, and neurophysiology perhaps due to the complexity ofthe cognition-action nexus and the lack of clear methodologicalapproaches within psychology (Rosenbaum, 2005).“Thinking andaction seem to lie at opposite ends of the behavioral spectrum”(Moran, 2012b, p. 1). The disembodied approach of informationprocessing theorists in the 1970s led scientists to conduct researchon thinking independently from the study of sensorimotor pro-cesses and mechanisms (Laakso, 2011). It was not until the adventof the motor cognition paradigm (Jeannerod, 1994) that “action”became subject to intensive scrutiny by researchers in psychology(MacIntyre et al., 2013). Jeannerod proposed that action, ratherthan movement per se, was vital to understand, as evidence forthe role of cognition in movement planning was accumulating.The interest in understanding action from different perspectiveswas increasing rapidly (see Guillot and Collet, 2010). Accordingto Moran (2009) importance of inter-disciplinary collaborationbetween researchers in cognitive sport psychology, cognitive psy-chology, and cognitive neuroscience has been brought to thefore by this new paradigm. Similarly, social cognition has devel-oped as a field of study which has added considerably to ourunderstanding of action (Gallese et al., 2004; Frith, 2012). Andrecently, cognitive researchers have embraced the study of thedomain of sport in their quest to understand how the mindworks (MacIntyre et al., 2013). Consequently, we propose thatmetacognition, a construct that is central to motor cognition,social cognition and action, can augment our current explanationsand understanding of the preparation and execution of motorskills within the sport context and elucidate our conceptions ofexpertise.

WHAT IS METACOGNITION?Metacognition, or “knowledge or cognition about cognitive phe-nomena” (Flavell, 1979, p. 906) is curiously under-explored inthe domain of expertise among sports performers (Moran, 1996;MacIntyre and Moran, 2010). Elite athletes are not just expertsin movement execution but conceivably they are also experts inplanning, metacognition, and reflection. Metacognition has beendefined as an individual’s insight into and control over their ownmental processes (Flavell, 1979). In the decades since Flavell’s(1979) pioneering article, the term metacognition has been oper-ationalized as the scientific study of the mind’s ability to monitorand control itself or, in other words, the study of our ability toknow about our knowing (Van Overschelde, 2008, p. 47). It is adifferent kind of cognition as explained by Nelson (1999, p. 625):“If one aspect of cognition is monitoring or controlling another

aspect of cognition, then the former aspect is metacognitive inrelation to the latter aspect. Flavell and subsequent investiga-tors have suggested a tripartite model of metacognition, withknowledge, control and monitoring components (Flavell, 1979,1987; Tarricone, 2011; Halpern, 2014). Recently, Tarricone (2011)indicated that the main interaction between metacognition andself-regulation is to control, monitor, and regulate strategies tomeet task demands and goals. Previously, the study of metacog-nition has targeted intellectual skills and a substantial corpus ofresearch exists on metacognition in educational settings (Hackeret al., 2009). However, Augustyn and Rosenbaum (2005, p. 911)recently challenged the status quo in metacognition research andstated that “if intellectual and perceptual-motor skills rely on sim-ilar mechanisms, one would expect metacognition to apply to theguidance of perceptual-motor skills, just as it does to the guidanceof intellectual skills.” The approach among cognitive neuroscien-tists, focused on visual perceptual tasks to measure metacognition(Palmer et al., 2013; Weil et al., 2013), is similarly narrow, per-haps due to methodological issues. Metacognition and action,on the other hand, offers new possibilities in illuminating ourunderstanding of expertise and action.

EXPERTISE AND METACOGNITIONExpertise is tightly coupled with metacognition in both training(e.g., knowledge of when a skill has been acquired) and com-petitive settings (e.g., self-regulation under stress). We proposethat metacognitive processes are inherently related to expertise insports and we have summarized recent findings in the sport liter-ature that reflect the prominence of metacognitive explanations(see Table 1). Early investigations were focused on judgmentsof learning (Simon and Bjork, 2001) and more recently morespecificity in the research questions has led to the developmentof specific models (MacIntyre and Moran, 2010). In the comingsections, we postulate that people use different sources of infor-mation, including metacognitive inferences. Firstly, we contendthat expertise in any given area facilitates metacognitive inferenceand secondly, that expertise itself may consist of metacognitiveinference, among a range of other non-metacognitive processesincluding working memory and motivation. Given that exper-tise is explained by differences in knowledge, many processesinvolving the use of that knowledge are more or less auto-matic or procedularized, and consequently they do may notplace onerous demands on working memory (Beilock and Carr,2001). This creates the opportunity for metacognitive reason-ing to optimize the assessment of situations and to structureone’s goal pursuit. Furthermore, experts have quite good ideasabout standards and deviations from such standards, whetherthis refers to one’s own behavior or to the behavior of oth-ers. Deviations from sophisticated mental models (e.g., theideal long jump) are thus more likely to become salient toexperts than to non-experts, also providing opportunities forreflective thoughts and interventions. The use of both actionsimulations (e.g., mental practice) and pre-performance rou-tines by elite performers can be conceptualized as domain-generalstrategies which rely upon metacognitive processes. Evidence tosupport these contentions will be presented in the forthcomingsections.





Table 1 | Recent research in sport that has highlighted the role of metacognition.

Topic Authors Method Emerging literature on metacognition

Meta-imagery Moran (2002) Conceptual Posed the question of whether meta-imagery abilities would

distinguish elite from non-elite performers.

Attentional strategies Nietfield (2003) Survey Indicated that the runners relied mostly on monitoring and

information management strategy (i.e., strategy thoughts during

running) regulatory cognitions.

Meta-imagery Weinberg et al. (2003) Survey Athletes were surveyed, not just on their use of imagery, but on

their perceived effectiveness of imagery for distinct functions.

Meta-imagery MacIntyre and Moran (2007a) Qualitative Relative to non-athletes (Denis and Carfantan, 1985), the participants

demonstrated an intricate understanding of imagery processes

including the use of imagery of realistic behaviors as opposed to

ideal performance imagery.

Meta-imagery MacIntyre and Moran (2007b) Qualitative Evidence from elite performers suggested that they possessed

sophisticated meta-imagery control skills – being able, for instance,

to restructure negative imagery so that it facilitates future

performance.

Psychological skills training Foster and Weigand (2008) Conceptual Psychological skills in sport (e.g., imagery, goal setting) can be

applied more efficiently, particularly in developmental contexts, by

applying metacognitive models to understand the role of

self-monitoring and self-regulation in the application of the above

strategies.

Meta-imagery MacIntyre and Moran (2010) Conceptual A model of meta-imagery was proposed with a specific emphasis on

expertise effects.

Attention and choking behavior DeCaro et al. (2011) Experimental Skill failure was linked to the extent to which skill execution depends

on explicit attentional control. Increased metacognitive awareness

may cause performers to evaluate their performance diverting

attention away from skill execution.

Meta-imagery Pearson et al. (2011) Experimental Findings support the role of meta-cognitive knowledge of imagery

ability and relate it to our ability to judge individual episodes of

imagery.

Attention and ironic processes Toner et al. (2013) Experimental Over-compensatory behavior was more prevalent amongst

low-skilled than high-skilled golfers and they concluded that future

research explore metacognition.

Attentional strategies Brick et al. (2014) Conceptual The authors in developing a tentative framework for attentional focus

in endurance activity, highlighted the potential benefits of applying a

metacognitive approach in future studies.

THE ROLE OF METACOGNITIVE INFERENCESThe Coliseum, Los Angeles, XXIII Olympic Games, August 8th1984. A strong wind was swirling around the stadium in theafternoon as Carl Lewis’s was preparing for his long jump (50stunning Olympic moments No. 44: Carl Lewis’s four golds in1984, 2012). ABC network commentators were referring to CarlLewis’ adjustments: “He has to block that out and has to only thinkabout heading down the runway and getting off as long a jump aspossible.” The other reporter then stated: “He has got a bit of aheadwind. I think that’s what he was waiting for. . . to decide whathe is going to do with his step with regard to this wind.” Carl Lewis

ran down the runway and leapt into the history books with a jumpthat at that stage was 50 cm greater than his rivals. Nevertheless,the commentators stated that it“looked like a very restrained effortto me, the last four or five strides. . . he really looked like he wassort of in that same stride that he was running in the last 50 m ofthe 200 this morning [200 m heats]” (Corry, 1984).

After fouling his next jump, Lewis decided not to take his fourother allotted jumps and many of the crowd responded by booingdespite the margin of victory ultimately being 30 cm (Corry, 1984).Many of the spectators plausibly wanted him to break Bob Bea-mon’s longstanding world record. However, Lewis’ rationale was





that he had other goals to achieve (e.g., to equal Jesse Owens fourtrack and field gold medals in 1936) and he still had his additionalrounds to run in the 200 and 100 m relay. Given that Carl Lewiswon four gold medals, just as Jessie Owens had, we can concludethat his strategies were successful. Both, the execution of the finaljump, in addition to his decision to rest, indicate metacognitiveinferences in addition to his athletic expertise.

When individuals perform skills in the sporting context thesocial situation provides many pieces of information. For exam-ple, when Carl Lewis is about to perform a long jump, he mayremember his coach’s words about his holding back on his speedboth to hit the board accurately and to preserve energy. As helooks down the runway and notices that it is somewhat slippery,and is unsure whether his left foot will begin to ache again, justas it did 30 years ago he is metacognitively engaged. There maybe even much more metacognitive activity occurring than theaforementioned examples may suggest. The cognitive system ischallenged with the amount of information that is accessible at agiven time and the implications that these pieces of informationmay have for the execution of skill need to be considered. Foroptimum adjustments, people need to be attuned to thoughts andfeelings about the self and the requirements of the social situa-tion. Yet, given that most social situations provide a tremendousamount of information, people need to be able to focus, that is,they need to be selective. So, when it comes to particular taskssuch as a long jump, people need to select “what is relevant” andignore “what is not relevant” for the task at hand. How does thiswork?

A model of metacognitionBless et al. (2009) proposed a model of metacognition that focuseson different information types. Specifically, cognitively accessi-ble declarative knowledge (i.e., knowledge about something) andfeelings (e.g., how people feel about an action) and memories(e.g., whether they can remember a coach’s instruction) serveas information about the situation and the target in question.This has particular resonance when people are uncertain aboutthe judgmental target or when affective states are not in line withimplications of the situation, then metacognitive processes aremore likely to come into play (e.g., Martin et al., 1993; Bless et al.,2009). Metacognitive inferences are governed by rules or theoriesthat decide whether the accessible information should be used andhow it should be used in the moment.

When an expert judges performance, this may access a rep-resentation about the target behavior (e.g., a motor image). Forexample, imagine Carl Lewis watching his 1984 long jumps atthe Olympic Games. A vast amount of information about longjumps, such as Bob Beamon’s Leap of the Century at the OlympicGames in Mexico City in 1968, and his own jump might be cog-nitively accessible at the time. Which pieces of information arerelevant? People apply filtering rules, considering whether theinformation is relevant and representative for the target behav-ior (e.g., Martin, 1986; Bless et al., 2009). Information mightjust as easily become accessible in a conversation and thus com-prise declarative information. Declarative knowledge is definedas “knowing that” (e.g., taping factual information) and contrastswith procedural knowledge, or skill memory, which is explained

as “knowing how” (Sternberg and Sternberg, 2012). For exam-ple, Carl Lewis might have a chat with Bob Beamon and MikePowell about his 1984 long jump performance. If informationbecomes accessible as part of conversations, the person also needsto ask him or herself whether the information is appropriate to beused for a judgment or decision. Here conversational rules as out-lined seem to guide decisions on relevance and informativeness ofcommunicated information (e.g., Schwarz, 1994; Igou and Bless,2003, 2005, 2007; Bless et al., 2009). Furthermore, metacognitionsmay relate to past or future affective states. These are thoughtsthat include many different sources of information (e.g., Wil-son and Gilbert, 2003; Igou, 2004, 2008) and assessments of thepast or forecasts of future affective states are likely to influenceone’s behavior (Kahneman and Snell, 1992). For example, a longjumper may know that he has felt good being in competitionsand thus decides to take part in one that is coming up in 2 weekstime.

Another important type of information and likely source ofmetacognitive processes are people’s feelings at the time of judg-ments or behavior. For example, a long jumper’s mood, emotion,or the ease with which information comes to mind, may influ-ence the execution of the long jump. Importantly, these feelingshave informational value (e.g., Schwarz, 2002) for the judgmentor behavior at hand. Usually, affective feelings are distinguishedfrom cognitive feelings (Bless et al., 2009). Moods and emotionsare considered affective feelings, and they can inform us aboutthe situation and the target (e.g., Schwarz, 1990). For example,according to Schwarz and Clore (1983), when asked to judge atarget (e.g., life in general), people ask themselves “How do I feelabout it?” which leads to positive evaluations when people are ina positive mood, and to negative evaluation when they are in anegative mood (e.g., Forgas, 2001).

The feeling of knowing (e.g., Koriat, 1993), ease of retrieval(e.g., Schwarz et al., 1991), familiarity (e.g., Jacoby et al., 1989),and processing fluency (e.g., Reber et al., 1998) are examples ofcognitive feelings (e.g., Bless et al., 2009; Huntsinger and Clore,2011). In a nutshell, these are all experiences that accompany cog-nitive processes and interact with these processes by serving asinformation about judgmental targets. For example, according tothe ease of retrieval heuristic (Schwarz et al., 1991), if it is easy tothink of having performed long jumps, then one would be likelyto evaluate one’s jumping capacity more favorably, than if it wasdifficult to retrieve such examples.

Memories have direct effects on how cognitive representationsare formed. However, congruent with Strack and Bless (1994)and Bless et al. (2009), we argue that people also use theoriesabout the functioning of memory as an indicator as to whetheraccessible information is valid and relevant for a judgment athand. For example, Carl Lewis may not remember that the coachwarned him to run within himself for the long jump run-up. Pos-sibly, Carl would reason that he would remember that becauseit would have been very untypical for his coach to hold thisopinion.

Our conceptualization of metacognition is in line with a broaddefinition of the construct, namely that metacognition is cognitionabout both thoughts and feelings. However, the narrower defini-tion of metacognition as control process (e.g., Shea et al., 2014) is





also just as valid. The latter refers to processes in which attentionand cognitive control are essential in structuring peoples’ thoughtsand actions. As Shea et al. (2014) describe in detail, these types ofthoughts are associated with cognitive effort and limited capac-ity (cognitive system 2) rather than automatic, effortless processes(cognitive system 1; Stanovich, 2011).

Generally, monitoring and excluding accessible items of infor-mation requires cognitive resources. As a result, reduction incognitive capacity increases the likelihood that relatively irrele-vant information is used for the judgment task at hand (Blesset al., 2009). To be clear, we do not think that metacognitionsalways need awareness and controlled thought processes; however,we believe that metacognitive inferences are especially influentialwhen people need to engage in this type of reflective behaviorin order make a judgment or decision. This is especially the casewhen situations are ambiguous and complex, when more or lessautomatic processes fail, or when accessible information is con-flicting, contradictory, or perceived as inappropriate for the taskat hand.

PST AND METACOGNITION: THE CASE FOR DOMAIN-GENERAL SKILLSInterestingly, the ability of Carl Lewis to combine excellence inboth track (e.g., 100, 200 m sprints and relay) and field events(e.g., long jump) supports the domain-specificity of expertise. Toexplain, long jump performance is determined largely by the ath-lete’s ability to attain a fast horizontal speed at the end of theapproach runway, thus the physiological task demands were com-patible (Bridgett and Linthorne, 2006). However, it is also likelythat what are termed “domain-general skills” including psycho-logical skills and metacognition may have played a role in his nineOlympic gold medal winning performances.

As noted earlier (see Table 1), Foster and Weigand (2008) high-lighted that some theoretical inadequacies in sport psychologycould be reconciled by considering other conceptual frameworks,including meta-cognitive approaches (Flavell, 1979). For exam-ple, by augmenting our understanding of psychological skills insport with the construct of metacognition, we could more clearlyunderstand the role of self-monitoring and self-regulation in theapplication of the aforementioned strategies.

Moran (1996) suggested that PST in sport is essentially an exer-cise in meta-cognitive instruction. Thus, in order to help athletesbecome independent thinkers, we need to know what they knowand believe about how their own minds work. In this regard, meta-cognitive control processes are especially valuable because theyallow people to change their behavior strategically in accordancewith task demands. Eccles and Feltovich (2008) proposed thataccelerated learning and enhanced performance, and ultimatelyexpertise, may be the result of a combination of psychologicalsupport skills (e.g., self-talk, goal-setting, relaxation, and mentalpractice) and metacognitive abilities. They are domain-general inthat they “can be applied to a variety of novel tasks and domains”(Eccles and Feltovich, 2008, p. 43). Meta-cognitive skills in thiscase are higher order skills that regulate learning and performance,including the coordination of the use of psychological supportskills (i.e., PST).

Within sport psychology, psychological skills have been shownto differentiate successful Olympians from their less successful

counterparts (Orlick and Partington, 1988; Gould et al., 2002;Fletcher and Sarkar, 2012). The coordinating role of metacogni-tion may be a key factor in the efficient use of psychological skills.Furthermore, emotional regulation is trainable and sustainable bythe application of PST and this has applications beyond the realmof sport (Eccles et al., 2011). Two other aspects of psychologicalskills that are indeed trainable are now discussed, meta-imageryand routines.

Is meta-imagery linked to expertise?One dimension of metacognition that has been illuminated byrecent research activity is “meta-imagery,” a performer “beliefsabout the nature and regulation of their own imagery skills”(Moran, 2002, p. 415). In 2002, little was known about thistopic relative to the knowledge base on other aspects of imagery,such as motor imagery. Over the preceding years research inthe expertise literature emerged to suggest that meta-imagery isanother factor that differentiates novices from experts (Moranet al., 2012). Researchers had explored the topic by asking athletesto indicate why, where, how, what, and when they use men-tal imagery processes (e.g., Munroe et al., 2000; MacIntyre andMoran, 2007a,b). Athletes’ responses from both interviews andsurveys demonstrated a comprehensive knowledge of the multi-modal potential of imagery, showed they employed imagery increative ways for contingency planning (see Moran, 2009) andwere also aware of robust imagery effects (e.g., mental prac-tice). Interestingly, a meta-analysis conducted in 1994, indicateda possible constraint on the efficacy of mental practice for novicelearners (i.e., experts improved more). Driskell et al. (1994) sug-gested that novices may not have an appropriate approximationof the motor skill or that their imagery abilities may be insuffi-cient to generate and manipulate the requisite visuo-spatial motorconfiguration. An alternative possibility is that experts may sim-ply possess greater meta-cognitive knowledge of how to employimagery effectively for skill improvement as compared to novices(MacIntyre et al., 2013). In fact, a model of meta-imagery wasdeveloped to account for the above findings and this also gener-ates possibilities of developing a test of meta-imagery (MacIntyreand Moran, 2010).

Furthermore, contemporary evidence from cognitive psychol-ogy supports the role of meta-cognitive knowledge of imageryability and relates it to our ability to judge individual episodes ofimagery (Pearson et al., 2011). The voluntary nature of imageryand the role of conscious awareness during imagery tasks makeit amenable to introspection, ironically the method that was cen-tral to the demise of the scientific study of imagery, a century ago(Roeckelein, 2004).

Is winning just a matter of routine?Pre-performance routines are integral to performance excellencein many self-paced sporting skills, from sprint running to penaltytaking in field games (Singer, 2000, 2002; Jackson and Baker, 2001).Defined by Moran (1996, p. 177) as “a sequence of task-relevantthoughts and actions which an athlete engages in systematicallyprior to his or her performance of a specific sports skill.” Thewidespread use of routines in sport demonstrates that attention iscentral to cognitive sport psychology because the ability to exert





mental effort effectively is vital for optimal athletic performance(Moran, 2009). One function of routines is to regulate arousalprior to skill execution and evidence for their role in bufferingstress or choking has also emerged (Mesagno and Mullane-Grant,2010). While routines have been explored across a range of sports,perhaps, the sport of golf has received the most attention fromresearchers (e.g., McCann et al., 2001; Cotterill et al., 2010). Inter-estingly, three time major winner golfer Padraig Harrington isquoted as saying “my key isn’t working at the moment so wehave to figure out a way. . . I have gone a bit stale focusingon the target.” What is instructive about this statement is thatthe golfer appears to realize that his routine is no longer func-tioning appropriately. Routines need to be revised regularly toavoid the routine itself becoming too automatic (Moran, 1996).From a metacognitive perspective, this may be accounted for bymetacognitive monitoring. Thus metacognition may be funda-mental to the refinement of pre-performance routines as wellas their acquisition. A recent review noted that “at a funda-mental level it is still not clear what function routines fulfill,what they should consist of or the most effective way to teachthem” (Cotterill, 2010, p. 132). The potential for metacogniton

research to shed light on the development and refinement ofroutines as well as their theoretical and conceptual basis is readilyapparent.

NEW AVENUES FOR FUTURE RESEARCHIn the preceding paragraphs, arguments for the potential for theconstruct of metacognition to clarify our understanding of exper-tise have been made. Now we wish to specify research topicsaugmented by appropriate methodologies and possible tasks (seeTable 2).

MeasurementOne of the key challenges in the operationalization of any con-struct is the development of appropriate measurement tools. Atpresent, there is a paucity of questionnaires to assess metacog-nitive abilities. One such measure is the 52-item MetacognitiveAwareness Inventory (Schraw and Dennison, 1994) which employsa two factor model (knowledge of cognition; regulation of cog-nition). Currently, there is a need to develop and validate revisedpsychometric instruments to assess, for example, meta-imagerybeliefs and knowledge. Further questionnaires for meta-attention

Table 2 | Proposed research topics, methods, and objectives of future studies to study expertise and metacognition in sport settings.

Research topic Methods Objectives

Measurement Psychometric and

experimental approach

To engage in conceptual analysis of the construct of metacognition (and related constructs).

To investigate, using dual-task methods, the distinct cognitive processes underlying metacognition (e.g.,

interference with aspects of working memory).

To create psychometrically valid measures of metacognition and action processes (e.g., meta-imagery).

Motor cognition Action simulation and

converging methods

To generate specific hypotheses to empirically test if metacognition is a domain-general skill across

different motor simulation processes.

To use paradigms from motor cognition (e.g., mental travel studies) to evaluate metacognitive monitoring

ability.

Anxiety Experimental and field

study approach

To investigate stereotype threat and the interaction with metacognitive processes in both well-learned

and novel skills.

To examine how current affect and anticipated affective responses to performance influence action in

sports via meta-cognitive thoughts.

To examine how metacognitive training can influence skilled performance and athletes’ susceptibility to

overcompensation of attention.

Neuroscience Neural imaging To investigate the neural substrates of the existing models of cognitive control that relate to

metacognition processes.

To specify the neural architecture underlying metacognitive abilities.

To elucidate whether metacognition is linked to a global mechanism or if distributed neural substrates

underlie different components of metacognition.

Developmental Mixed-methods To assess the role of cognitive development in the acquisition of meta-cognitive skills.

To explore the potential of interventions to enhance metacognitive abilities among those who experience

deficits in, for example, their judgments of learning.

To understand the interaction between metacognitive abilities and motor skill acquisition across the

lifespan (e.g., how the elderly can cope using metacognitive skills to supplement diminishing working

memory).





or social metacognition (for team sports) could also be piloted,refined and analyzed using factor analysis. Parallel with the objec-tive would be the refinement of the construct of metacognitionas it relates to both expertises. Tarricone (2011) had conducteda comprehensive “taxonomy of metacognition” which is primar-ily focused upon on the wide-scale research in the educationalresearch domain. A similar task would be beneficial for metacog-nition within the context of expertise and the models discussedin this review (e.g., Bless et al., 2009). The approach of Flemingand Lau (2014, p. 1) who distinguish between “metacognitive bias(a difference in subjective confidence despite basic task perfor-mance remaining constant), metacognitive sensitivity (how goodone is at distinguishing between one’s own correct and incor-rect judgments) and metacognitive efficiency (a subject’s levelof metacognitive sensitivity given a certain level of task perfor-mance)”raises interesting questions for the study of metacognitionand conscious awareness. Their approach, focusing on percep-tual expertise, includes elements which have a direct relevanceto expert performance. For instance, they suggest that “metacog-nitive confidence” can be interpreted as a probability judgmentdirected toward one’s own decisions-the probability of a previousjudgment being correct. This is synonymous with expertise (i.e.,ability to predict actions). It is our view that the exploration of theconstruct of metacognition and how they interface with actionrelated processes (e.g., motor imagery), we can contribute to theconceptual development of the construct of metacognition. Thedivergent approaches to date necessitate a degree of conceptualanalysis, a process that is all too rare in psychology (Machado andSilva, 2007).

Motor cognitionRecent conceptualizations of imagery, action observation andmotor execution, view these processes as overlapping, differingby degree rather than kind (Jeannerod, 1994, 2006; Vogt et al.,2013). The preceding section on meta-imagery is illustrative ofthe progress that can be made in our understanding of exper-tise, metacognition and imagery, alike. Given the overlap betweenfor example, imagery (e.g., visualizing a long jump-the run-up,take-off, and landing phases) and action observation (e.g., view-ing Bob Beamons’ world record long jump), a question arises as towhether the same metacognitive processes underlie these relatedprocesses. This issue is further complicated by evidence from sev-eral sources which suggests motor imagery is grounded in physicalexperience, for example, the specific training either simulated orexecuted (Olson and Nyberg, 2011; Debarnot et al., 2014). Thequestion remains as to whether the respective metacognitive pro-cesses are domain-general or domain-specific? As a result, deeperconceptual analysis is required to comprehensively describe andexplain the range of metacognitive processes that pertain to cog-nitive simulation strategies. This new dimension to metacognitionresearch offers a range of experimental possibilities that can enablea greater understanding of metacognition with regard to actionpreparation, simulation, and execution.

AnxietyResearch on “choking” in sport has illuminated our understand-ing of anxiety across both cognitive skills (e.g., Lyons and Beilock,

2012) and motor skill contexts (e.g., Beilock and Carr, 2001;Beilock and Gonso, 2008; DeCaro et al., 2011; Toner and Moran,2011; Toner et al., 2013). The “explicit monitoring hypothesis”suggests that attending to a well learned skill may lead to failurein the precise execution of the skill under pressure. Metacogni-tive abilities obviously have a role in regulation emotion, based onthe aforementioned model by Bless et al. (2009). Furthermore,the role of “stereotype threat” which occurs when “knowledgeof a negative stereotype about a social group leads to a les-than-optimal performance by members of that group” shouldbe investigated from a metacognitive perspective (Beilock andMcConnell, 2004). Previous investigations supported the con-tention that stereotype threat prompts attention to the executedaction and thus can disrupt performance (Beilock et al., 2003).However, this effect can be alleviated by the inclusion of a sec-ondary task. Findings across two studies conducted by Beilocket al. (2003) were inconsistent and the impact of stereotype threatmay be more telling across a tournament than a putting skillas it may interfere with the metacognitive processes that helpmodulate attention and regulate emotion. Another avenue is tosystematically examine how current affect and anticipated affectiveresponses to performance influence action in sports via meta-cognitive thoughts. This is based on the recent literature onaffect regulation (e.g., Baumeister et al., 2007; Loewenstein, 2011)Moreover, it would be interesting to investigate how trainingin metacognition can influence skilled performance and ath-letes’ susceptibility to overcompensation of attention and affectregulation.

NeuroscienceThe rise of neuroscience in recent decades has been based largelyupon advances in methodologies that facilitate the study of inter-nal mental experiences, such as metacognition, in a robust andscientific way. Cognitive neuroscience, in particular, has had adramatic effect on our understanding of individual domains ofcognition from vision to memory (Beran et al., 2012), in chess(Bilalic et al., 2010, 2012; Bartlett et al., 2013) and more recentlyin sport (Debarnot et al., 2014). As we have seen throughout thecurrent paper sport and athletic skills offer a dynamic and fascinat-ing arena to study and explore cognitive processes, metacognitiveprocesses, and experiences.

Regarding expertise, we saw that in order to engage in effec-tive training episodes for long periods of time, athletes must behighly self-disciplined and self-regulated (Crews et al., 2001). Thisnotion of self-regulation, defined as a set of cognitive, behavioral,and motivational processes that interact to influence performance(Kitsantas and Kavussanu, 2011) has been the go-to approachfor examining expertise differences in performance domains.This approach has been concerned with self-regulatory processes(imagery, attentional control, for example) and researchers typi-cally attempted to make confidence judgments about the efficacyof some aspect of their cognition. Neuroscience has enabledresearchers to move beyond the study of processes and focus onmetacognitive judgments instead (a case in point being; the feel-ing of inaccessibility otherwise known as the tip of the tonguephenomena). This temporary failure of retrieval for a memoryhighlights a problem with a particular cognitive process but not





a problem with one’s metacognitive judgment. What tip of thetongue research has shown us is that different underlying pro-cesses are responsible for the cognition and metacognition thatmonitors it (Schwartz and Diaz, 2014). Metacognitive experiencesarise from cognitive processes and correspond to particular behav-iors. The cognitive processes that produced the behavior are notthe same as the processes that gave rise to the metacognition.For example, an object is recognized as having been seen before(cognitive process) accompanied by an experience of confidence,and the person then says that they know the answer (behavior).Essentially, there are a set of cognitive processes driving the recallof information but another set of processes driving our aware-ness of it. “Thus understanding any metacognitive judgment mustinvolve understanding the cognition it measures and the multipleprocesses that contribute to the judgment” (Schwartz and Diaz,2014, p. 9). Attempts at componential analysis of metacognitionare in their infancy (Fleming and Lau, 2014; Garrison, 2014), butthey appear to be fruitful with regard to understanding its impactupon visual perceptual tasks. Nevertheless, the investigation of theneural basis of metacognition (Baird et al., 2013) is not withoutits limitations. It has been noted that the application of neu-rophysiological measurement techniques impose restrictions onthe ecological validity of studies which are not readily overcome(Mann et al., 2013).

DevelopmentalCurrently, a gap exists in our knowledge of how performersacquire pre-performance routines (Singer, 2000, 2002; Cotter-ill, 2010). Unfortunately, researchers have neglected to explorehow these strategies are developed over time with one recentnotable exception, a study with gymnastics athletes (Faggiani et al.,2012). The role of cognitive development in the acquisition ofmeta-cognitive skills may be a limiting factor for applied sportpsychology interventions (Foster and Weigand, 2008). Thus thegap in the knowledge base may be due to the complex inter-action of domain-general and domain-specific cognitive skills.Given that our approach has centered on the role of metacog-nitive abilities and processes, we propose that a developmentalapproach to understanding pre-performance routines could beaugmented by exploring metacognitive skill development from alongitudinal perspective. The potential of specific interventionsto enhance metacognitive ability could be explored for those whoare impaired in their metacognitive development or for those whosuffer plateaus in their skill development. For example, recentresearch has demonstrated the ability of a 2-week meditation pro-gram in enhancing metacognitive ability in a perceptual task (Bairdet al., 2014).

CONCLUSIONDo metacognitive processes enhance performance? Are they help-ful in the acquisition of expertise? The general answer to thesequestions is in the affirmative. Metacognitive processes are part ofthe inventory of human thought (Nelson and Narens, 1994; Stern-berg, 2001; Perfect and Schwartz, 2002). As such, they serve as aresource to structure thought and regulate behavior. Will metacog-nitive processes always lead to better outcomes? No, certainlynot. As for all areas of information processing, people can err.

However, understanding the role of metacognition, the breadthand flexibility of processes involved and how they are associatedwith expertise, allows for more precise predictions of behavior.

We argue that more research and empirical scrutiny of the con-struct of metacognition can help to develop principles that governthe relation between internal cognitive processes and subjectiveexperience. These principles could be very effective for expertiseresearch looking to differentiate a “real” expert from a “skilled”performer, currently a major challenge in expertise research (e.g.,MacIntyre et al., 2013; Bourne et al., 2014).

In the sporting domain, training athletes to initiate, developand engage in metacognition can equip them with the properstrategies, beliefs, and self-understanding to excel in sports.Currently, there is no unified view as to what athletic train-ing entails and coaches, despite a burgeoning literature (Healyet al., 2014) have, for example, focused on a relatively nar-row set of conclusions from the deliberate practice literature(Baker and Young, 2014). Metacognition offers the potential toexpand our understanding of expertise and individual domainsof cognition through a rigorous examination of the mecha-nisms underlying self-initiated monitoring and control of onesown performance. Consequently, our understanding of expertisecan be illuminated by studying metacognition in the sporting,domain, specifically, by using a strength-based approach withexpert samples (MacIntyre et al., 2013). Sport offers researchersa fertile natural laboratory where expertise is easily quantifiablethrough the quest to be faster, higher and stronger. In conclu-sion, we have demonstrated that the construct of metacognitionhas the potential to be a springboard for research into sportingexpertise.

ACKNOWLEDGMENTSThis work was supported jointly by the Irish Research Coun-cil under their ‘New Foundations’ scheme and the Departmentof Physical Education and Sport Science, University of Limerick‘Start-Up’ grant scheme awards to the first author. We wish toacknowledge the efforts of the three reviewers and editorial teamin refining our ideas and assisting us in creating a clearer narrative.

REFERENCES50 stunning Olympic moments No. 44: Carl Lewis’s four golds in 1984. (2012). The

Guardian. Available at: http://www.theguardian.com [accessed June 29, 2012].Augustyn, J. S., and Rosenbaum, D. A. (2005). Metacognitive control of action:

preparation for aiming reflects knowledge of Fitts’ law. Psychon. Bull. Rev. 12,911–916. doi: 10.3758/BF03196785

Baird, B., Mrazek, M. D., Phillips, D. T., and Schooler, J. W. (2014). Domain-specific enhancement of metacognitive ability following meditation raining. J.Exp. Psychol. Gen. 143, 1972–1979. doi: 10.1037/a0036882

Baird, B., Smallwood, J., Gorgolewski, K. J., and Margulies, D. S. (2013).Medial and lateral networks in anterior prefrontal cortex support metacogni-tive ability for memory and perception. J. Neurosci. 33, 16657–16665. doi:10.1523/JNEUROSCI.0786-13.2013

Baker, J., and Young, B. (2014). 20 years later: deliberate practice and the devel-opment of expertise in sport. Int. Rev. Sport Exerc. Psychol. 135–157. doi:10.1080/1750984x.2014.896024

Bartlett, J. C., Boggan, A. L., and Krawczyk, D. C. (2013). Expertise andprocessing distorted structure in chess. Front. Hum. Neurosci. 7:825. doi:10.3389/fnhum.2013.00825

Baumeister, R. F., Zell, A. L., and Tice, D. M. (2007). “How emotions facilitate andimpair self-regulation,” in Handbook of Emotion Regulation, ed. J. J. Gross (NewYork, NY: Guilford Press), 408–428.


http://www.theguardian.com




Beilock, S. L., and Carr, T. H. (2001). On the fragility of skilled performance:what governs choking under pressure? J. Exp. Psychol. Gen. 130, 701–725. doi:10.1037/0096-3445.130.4.701

Beilock, S. L., and Gonso, S. (2008). Putting in the mind versus putting on the green:expertise, performance time, and the linking of imagery and action. Q. J. Exp.Psychol. 61, 920–932. doi: 10.1080/17470210701625626

Beilock, S. L., Jellison, W. A., McConnell, A. R., and Carr, T. H. (2003). Whennegative stereotypes lead to negative performance outcomes: understanding thestereotype threat phenomenon in sensorimotor skill execution. Paper presentedat the North American Society for the Psychology of Sport and Physical Activity.Savannah, GA.

Beilock, S. L., and McConnell, A. R. (2004). Stereotype threat and sport: can athleticperformance be threatened? J. Sport Exerc. Psychol. 26, 597–609.

Beran, M. J., Brandl, J., Perner, J., and Proust, J. (eds). (2012).Foundations of Metacognition. Oxford: Oxford University Press. doi:10.1093/acprof:oso/9780199646739.001.0001

Bilalic, M., Langner, R., Erb, M., and Grodd, W. (2010). Mechanisms and neuralbasis of object and pattern recognition: a study with chess experts. J. Exp. Psychol.Gen. 139, 728–742. doi: 10.1037/a0020756


Bless, H., Keller, J., and Igou, E. R. (2009). “Metacognition,” in Social Cognition,the Basis of Human Interaction. Frontiers of Social Psychology, eds J. Förster and F.Strack (New York: Psychology Press), 157–178.

Bourne, L. E., Kole, J. A., and Healy, A. F. (2014). Expertise: defined, described,explained. Front. Psychol. 5:186. doi: 10.3389/fpsyg.2014.00186

Brick, N., Mac Intyre, T., and Campbell, M. (2014). Attentional focus in enduranceactivity: new paradigms and future directions. Int. Rev. Sport Exerc. Psychol. 7,106–134. doi: 10.1080/1750984X.2014.885554

Bridgett, L. A., and Linthorne, N. P. (2006). Changes in long jump take-off technique with increasing run-up speed. J. Sports Sci. 24, 889–897. doi:10.1080/02640410500298040

Campitelli, G., and Gobet, F. (2005). The mind’s eye in blindfold chess. Eur. J. Cogn.Psychol. 17, 23–45. doi: 10.1080/09541440340000349

Campitelli, G., and Gobet, F. (2011). Deliberate practice: necessary but not sufficient.Curr. Dir. Psychol. Sci. 20, 280–285. doi: 10.1177/0963721411421922

Carson, H. J., and Collins, D. (2011). Refining and regaining skills in fixa-tion/diversification stage performers: the five-A model. Int. Rev. Sport Exerc.Psychol. 4, 146–167. doi: 10.1080/1750984X.2011.613682

Chase, W. G., and Simon, H. A. (1973). “The mind’s eye in chess,” in VisualInformation Processing, ed. W. G. Chase (New York: Academic Press), 215–272.

Corry, J. (1984). TV review; ABC’S coverage of the Olympics. The New York Times,11th February.

Cotterill, S. (2010). Pre-performance routines in sport: current understand-ing and future directions. Int. Rev. Sport Exerc. Psychol. 3, 132–153. doi:10.1080/1750984X.2010.488269

Cotterill, S. T., Sanders, R., and Collins, D. (2010). Developing effective pre-performance routines in Gglf: why don’t we ask the golfer? J. Appl. Sport Psychol.22, 51–64. doi: 10.1080/10413200903403216

Crews, D. J., Lochbaum, M. R., and Karoly, P. (2001). “Self-regulation: concepts,methods, and strategies in sport and exercise,” in Handbook of Sport Psychol-ogy, eds R. N. Singer, H. A. Hausenblas, and C. M. Janelle (New York: Wiley),566–581.

Debarnot, U., Sperduti, M., Di Rienzo, F., and Guillot, A. (2014). Experts bodies,experts minds: how physical and mental training shape the brain. Front. Hum.Neurosci. 8:280. doi: 10.3389/fnhum.2014.00280

DeCaro, M. S., Thomas, R. D., Albert, N. B., and Beilock, S. (2011). Choking underpressure: multiple routes to skill failure. J. Exp. Psychol. Gen. 140, 390–406. doi:10.1037/a0023466

de Groot, A. D. (1965). Thought and Choice in Chess, 2nd Edn. The Hague: Mouton.Denis, M., and Carfantan, M. (1985). People’s knowledge about images. Cognition

20, 49–60. doi: 10.1016/0010-0277(85)90004-6Detterman, D. K. (2014). Introduction to the intelligence special issue on

the development of expertise: is ability necessary? Intelligence 45, 1–5. doi:10.1016/j.intell.2014.02.004

Driskell, J. E., Copper, C., and Moran, A. (1994). Does mental practice enhanceperformance? J. Appl. Psychol. 79, 481–492. doi: 10.1037/0021-9010.79.4.481

Eccles, D. W., and Feltovich, P. J. (2008). Implications of domain-general “psycho-logical support skills” for transfer of skill and acquisition of expertise. Perform.Improv. Q. 21, 43–60. doi: 10.1002/piq.20014

Eccles, D. W., Ward, P., Woodman, T., Janelle, C. M., Le Scanff, C., Ehrlinger, J., et al.(2011). Human factors and sport psychology: advancing the science of humanperformance. Hum. Factors 53, 180–202. doi: 10.1177/0018720811403731

Ericsson, A. (2009). Development of Professional Expertise: Toward Measurementof Expert Performance and Design of Optimal Learning Environments. New York:Cambridge University Press. doi: 10.1017/CBO9780511609817

Ericsson, K. A. (2006a). “An introduction to Cambridge handbook of expertise andexpert performance: its development, organization, and content,” in CambridgeHandbook of Expertise and Expert Performance, eds K. A. Ericsson, N. Charness,P. Feltovich, and R. R. Hoffman (Cambridge: Cambridge University Press), 1–19.doi: 10.1017/CBO9780511816796

Ericsson, K. A. (2006b). “The influence of experience and deliberate practice onthe development of superior expert performance,” in Cambridge Handbook ofExpertise and Expert Performance, eds K. A. Ericsson, N. Charness, P. Feltovich,and R. R. Hoffman (Cambridge: Cambridge University Press), 685–706.

Ericsson, K. A. (2014). Why expert performance is special and cannot be extrap-olated from studies of performance in the general population: a response tocriticisms. Intelligence 45, 81–83. doi: 10.1016/j.intell.2013.12.001

Ericsson, K. A., Krampe, R. T., and Tesch-Romer, C. (1993). The role of deliberatepractice in the acquisition of expert performance. Psychol. Rev. 100, 363–406. doi:10.1037/0033-295X.100.3.363

Ericsson, K. A., and Smith, J. (1991). Toward a General Theory of Expertise: Prospectsand Limits. New York: Cambridge University Press.

Faggiani, F., McRobert, A. P., and Knowles, Z. (2012). Developing pre-performanceroutines for acrobatic gymnastics: a case study with a youth tumbling gymnast.Sci. Gymnast. J. 4, 39–52.

Flavell, J. H. (1979). Metacognition and cognitive monitoring: a new area ofcognitive-developmental inquiry. Am. Psychol. 34, 906–911. doi: 10.1037/0003-066X.34.10.906

Flavell, J. H. (1987). “Speculations about the nature and development of metacog-nition,” in Metacognition, Motivation, and Understanding, eds F. E. Weinert andR. H. Kluwe (Hillsdale, NJ: Lawrence Erlbaum Associates), 21–29.

Fleming, S. M., and Lau, H. C. (2014). How to measure metacognition. Front. Hum.Neurosci. 8:443. doi: 10.3389/fnhum.2014.00443

Fletcher, D., and Sarkar, M. (2012). A grounded theory of psychologicalresilience in Olympic champions. Psychol. Sport Exerc. 13, 669–678. doi:10.1016/j.psychsport.2012.04.007

Forgas, J. P. (2001). “The affect infusion model (AIM): an integrative theory ofmood effects on cognition and judgments,” in Theories of Mood and Cognition,eds L. L. Martin and G. L. Clore (Mahwah, NJ: Erlbaum), 99–134.

Foster, D., and Weigand, D. (2008). The role of cognitive and metacognitivedevelopment in mental skills training. Sport Exerc. Psychol. Rev. 4, 21–29.

Frith, C. D. (2012). The role of metacognition in human social interactions. Philos.Trans. R. Soc. B Biol. Sci. 367, 2213–2223. doi: 10.1098/rstb.2012.0123

Gallese, V., Keysers, C., and Rizzollatti, G. (2004). A unifying view of the basisof social cognition. Trends Cogn. Sci. 8, 396–409. doi: 10.1016/j.tics.2004.07.002

Garrison, J. (2014). Dissociable neural networks supporting metacogni-tion for memory and perception. J. Neurosci. 34, 2765–2767. doi:10.1523/JNEUROSCI.5232-13.2014

Gladwell, M. (2008). Outliers: The Story of Success. New York: Little, Brown andCompany.

Gobet, F., and Campitelli, G. (2007). The role of domain-specific practice,handedness, and starting age in chess. Dev. Psychol. 43, 159–172. doi:10.1037/00121649.43.1.159

Gould, D., Dieffenbach, K., and Moffett, A. (2002). Psychological characteristics andtheir development in Olympic champions. J. Appl. Sport Psychol. 14, 172–204. doi:10.1080/10413200290103482

Guillot, A., and Collet, C. (2010). The Neurophysiological Foundationsof Mental and Motor Imagery. Oxford: Oxford University Press. doi:10.1093/acprof:oso/9780199546251.001.0001

Hacker, D. J., Dunlosky, J., and Graesser, A. C. (2009). Handbook of Metacognitionin Education. New York, NY: Routledge.

Halpern, D. F. (2014). Thought and Knowledge: An Introduction to Critical Thinking,5th Edn. New York, NY: Psychology Press.





Hambrick, D. Z., Oswald, F. L., Altmann, E. M., Meinz, E. J., Gobet, F., andCampitelli, G. (2014a). Deliberate practice: is that all it takes to become anexpert? Intelligence 45, 34–45. doi: 10.1016/j.intell.2013.04.001

Hambrick, D. Z., Oswald, F. L., Altmann, E. M., Meinz, E. J., Gobet, F., andCampitelli, G. (2014b). Accounting for expert performance: the devil is in thedetails. Intelligence 45, 112–114. doi: 10.1016/j.intell.2014.01.007

Healy, A. F., Kole, J. A., and Bourne, L. E. (2014). Training principles to advanceexpertise. Front. Psychol. 5:131. doi: 10.3389/fpsyg.2014.00131

Helsen, W. F., Starkes, J. L., and Hodges, N. J. (1998). Team sports and the theory ofdeliberate practice. J. Sports Exerc. Psychol. 20, 13–35.

Hoffman, R. R. (2014). The Psychology of Expertise: Cognitive Research and EmpiricalAI. New York, NY: Springer.

Holding, D. H. (1985). The Psychology of Chess Skill. Hillsdale, NJ: ErlbaumAssociates.

Holding, D. H. (1992). Theories of chess skill. Psychol. Res. 54, 10–16. doi:10.1007/BF01359218

Huntsinger, J. R., and Clore, G. L. (2011). “Emotion and social metacognition,”in Social Metacognition, eds P. Briñol and K. DeMarree (New York: PsychologyPress), 199–217.

Igou, E. R. (2004). Lay theories in affective forecasting: the progression of affect. J.Exp. Soc. Psychol. 40, 528–534. doi: 10.1016/j.jesp.2003.09.004

Igou, E. R. (2008). “How long will I suffer?” versus “How long will you suffer?” Aself-other effect in affective forecasting. J. Pers. Soc. Psychol. 95, 899–917. doi:10.1037/a0011619

Igou, E. R., and Bless, H. (2003). Inferring the importance of arguments: order effectsand conversational Rules. J. Exp. Soc. Psychol. 39, 91–99. doi: 10.1016/S0022-1031(02)00509-7

Igou, E. R., and Bless, H. (2005). The conversational basis for the dilution effect. J.Lang. Soc. Psychol. 24, 25–35. doi: 10.1177/0261927X04273035

Igou, E. R., and Bless, H. (2007). Conversational expectations as a basisfor order effects in persuasion. J. Lang. Soc. Psychol. 26, 260–273. doi:10.1177/0261927X06303454

Jackson, R. C., and Baker, J. S. (2001). Routines, rituals, and rugby: case study of aworld class goal kicker. Sport Psychol. 15, 48–65.

Jacoby, L. L., Kelley, C. M., Brown, J., and Jaschenko, J. (1989). Becoming famousovernight: limits of the ability to avoid unconscious influences of the past. J. Pers.Soc. Psychol. 56, 326–338. doi: 10.1037/0022-3514.56.3.326

Jeannerod, M. (1994). The representing brain: neural correlates of motor intentionand imagery. Behav. Brain Sci. 17, 187–245. doi: 10.1017/S0140525X00034026

Jeannerod, M. (2006). Motor Cognition: What Actions Tells The Self. New York:Oxford University Press. doi: 10.1093/acprof:oso/9780198569657.001.0001

Kahneman, D., and Snell, J. (1992). Predicting a change in taste: do peo-ple know what they will like? J. Behav. Decis. Mak. 5, 187–200. doi:10.1002/bdm.3960050304

Kitsantas, A., and Kavussanu, M. (2011). “Acquisition of sport knowledge and skill:the role of self-regulatory processes,” in Handbook of Self-Regulation on Learningand Performance, eds B. J. Zimmerman and D. H. Schunk (New York: Routledge),217–233.

Klostermann, A., Kredel, R., and Hossner, E. (2013). The “quiet eye” and motorperformance: task demands matter! J. Exp. Psychol. Hum. Percept. Perform. 39,1270–1278. doi: 10.1037/a0031499

Koriat, A. (1993). How do we know what we do? The accessibility model of the feelingof knowing. Psychol. Rev. 100, 609–639. doi: 10.1037/0033-295X.100.4.609

Laakso, A. (2011). Embodiment and development in cognitive science. Cogn. BrainBehav. 4, 409–425.

Lewis, C., and Marx, J. (1996). One More Victory Lap: My Personal Diary of anOlympic Year. Santa Monica, CA: Athletics International.

Loewenstein, G. (2011). “Affect regulation and affective forecasting,” in Handbookof Emotion Regulation, ed. J. J. Gross (New York, NY: Guilford Press), 180–203.

Lyons, I. M., and Beilock, S. L. (2012). When math hurts: math anxiety predictspain network activation in anticipation of doing math. PLoS ONE 7:e48076. doi:10.1371/journal.pone.0048076

Machado, A., and Silva, F. J. (2007). Toward a richer view of the scientific method:the role of conceptual analysis. Am. Psychol. 62, 671–681. doi: 10.1037/0003-066X.62.7.671

MacIntyre, T., and Moran, A. (2007a). Exploring imagery use and meta-imageryprocesses: qualitative investigations with elite canoe-slalom athletes. J. ImageryRes. Sport Phys. Act. 2:3.

MacIntyre, T., and Moran, A. (2007b). Exploring imagery use and meta-imageryprocesses: qualitative investigations with an elite multi-sport sample. J. ImageryRes. Sport Phys. Act. 2:4.

MacIntyre, T., Moran, A., Guillot, A., and Collet, C. (2013). An emerging paradigm:a strength-based approach to exploring mental imagery. Front. Hum. Neurosci.7:104. doi: 10.3389/fnhum.2013.00104

MacIntyre, T., and Moran, A. P. (2010). “Meta-imagery processes among elite sportsperformers,” in The Neurophysiological Foundations of Mental and Motor Imagery,eds A. Guillot and C. Collet (Oxford: Oxford University Press), 227–244. doi:10.1093/acprof:oso/9780199546251.003.0016

Mann, D., Dicks, M., Cañal-Bruland, R., and van der Kamp, J. (2013). Neuro-physiological studies may provide a misleading picture of how perceptual-motorinteractions are coordinated. Iperception 4, 78–80. doi: 10.1068/i0569ic

Mann, D. T. Y., Williams, A. M., Ward, P., and Janelle, C. M. (2007). Perceptual-cognitive expertise in sport: a meta-analysis. J. Sport Exerc. Psychol. 29, 457–478.

Martin, L. L. (1986). Se/reset: the use and disuse of concepts in impres-sion formation. J. Pers. Soc. Psychol. 51, 93–120. doi: 10.1037/0022-3514.51.3.493

Martin, L. L., Ward, D. W., Achee, J. W., and Wyer, R. S. (1993). Mood as input:people have to interpret the motivational implications of their moods. J. Pers. Soc.Psychol. 64, 317–326. doi: 10.1037/0022-3514.64.3.317

McCann, P., Lavallee, D., and Lavallee, R. M. (2001). The effect of pre-shotroutines on golf wedge shot performance. Eur. J. Sport Sci. 1, 231–240. doi:10.1080/17461390100071503

Mesagno, C., and Mullane-Grant, T. (2010). A comparison of different pre-performance routines as possible choking interventions. J. Appl. Sport Psychol.22, 343–360. doi: 10.1080/10413200.2010.491780

Moran, A. (2009). Cognitive psychology in sport: progress and prospects. Psychol.Sport Exerc. 10, 420–426. doi: 10.1016/j.psychsport.2009.02.010

Moran, A., Guillot, A., MacIntyre, T., and Collet, C. (2012). Re-imagining mentalimagery: building bridges between cognitive neuroscience and sport psychology.Br. J. Psychol. 103, 224–247. doi: 10.1111/j.2044-8295.2011.02068.x

Moran, A. P. (1996). The Psychology of Concentration in Sport Performers: A CognitiveAnalysis. Hove, East Sussex: Psychology Press.

Moran, A. P. (2002). In the mind’s eye. Psychologist 15, 14–15.Moran, A. P. (2012a). Sport and Exercise Psychology: A Critical Introduction. London:

Routledge.Moran, A. P. (2012b). Thinking in action: some insights from cognitive sport

psychology. Think. Skills Creat. 7, 85–92. doi: 10.1016/j.tsc.2012.03.005Munroe, K. J., Giacobbi, P. R. Jr., Hall, C., and Weinberg, R. (2000). The four Ws of

imagery use: where, when, why, and what. Sport Psychol. 14, 119–137.Nelson, T. O. (1999). “Cognition versus metacognition,” in The Nature of Cognition,

ed. R. J. Sternberg (Cambridge, MA: The MIT Press), 625–641.Nelson, T. O., and Narens, L. (1994). Why Investigate Metacognition? Metacognition:

Knowing about Knowing. Cambridge, MA: The MIT Press.Nietfield, J. L. (2003). An examination of metacognitive strategy use and monitoring

skills by competitive middle distance runners. J. Appl. Sport Psychol. 15, 307–320.doi: 10.1080/714044199

Olson, C. J., and Nyberg, L. (2011). Brain simulation of action may be grounded inphysical experience. Neurocase 17, 501–505. doi: 10.1080/13554794.2010.547504

Orlick, T., and Partington, J. (1988). Mental links to excellence. Sport Psychol. 2,105–130.

Palmer, E. C., David, A. S., and Fleming, S. M. (2013). Effects of age on metacognitiveefficiency. Conscious. Cogn. 28, 151–160. doi: 10.1016/j.concog.2014.06.007

Pearson, J., Rademaker, R. L., and Tong, F. (2011). Evaluating the mind’seye: the metacognition of visual imagery. Psychol. Sci. 22, 1535–1542. doi:10.1177/0956797611417134

Perfect, T. J., and Schwartz, B. L. (2002). Applied Metacognition. Cambridge:Cambridge University Press. doi: 10.1017/CBO9780511489976

Reber, R., Winkielman, P., and Schwarz, N. (1998). Effects of perceptual fluency onaffective judgements. Psychol. Sci. 9, 45–48. doi: 10.1111/1467-9280.00008

Roeckelein, J. E. (2004). Imagery in Psychology: A Reference Guide. Westport, CT:Greenwood Publishing Group.

Rosenbaum, D. A. (2005). The cinderella of psychology: the neglect of motor con-trol in the science of mental life and behavior. Am. Psychol. 60, 308–317. doi:10.1037/0003-066X.60.4.308

Ross, P. E. (2006). The expert mind. Sci. Am. 295, 64–71. doi:10.1038/scientificamerican0806-64





Schraw, G., and Dennison, R. S. (1994). Assessing metacognitive awareness.Contemp. Educ. Psychol. 19, 460–475. doi: 10.1006/ceps.1994.1033

Schwarz, N. (1990). “Feeling as information: informational and motivational func-tions of affective states,” in Handbook of Motivation and Cognition: Foundationsof Social Behavior, eds R. M. Sorrentino and E. T. Higgins (New York: GuilfordPress), 527–561.

Schwarz, N. (1994). “Judgment in a social context: biases, shortcomings, and thelogic of conversation,” in Advances in Experimental Social Psychology, ed. P. Z.Mark (New York: Academic Press), 123–162.

Schwarz, N. (2002). “Situated cognition and the wisdom of feelings: cognitive tun-ing,” in The Wisdom in Feelings, eds L. F. Barrett and P. Salovey (New York: GuilfordPress), 144–166.

Schwarz, N., Bless, H., Strack, F., Klumpp, G., Rittenauer-Schatka, H., andSimons, A. (1991). Ease of retrieval as information: another look at the avail-ability heuristic. J. Pers. Soc. Psychol. 61, 95–202. doi: 10.1037/0022-3514.61.2.195

Schwarz, N., and Clore, G. L. (1983). Mood, misattribution, and judgments of well-being: informative and directive functions of affective states. J. Pers. Soc. Psychol.45, 513–523. doi: 10.1037/0022-3514.45.3.513

Schwartz, N., and Diaz, F. (2014). “Quantifying human metacognition for the neu-rosciences” in The Cognitive Neuroscience of Metacognition, eds S. M. Fleming andE. D. Frith (New York: Springer), 9–23. doi: 10.1007/978-3-642-45190-4_2

Shea, N., Boldt, A., Bang, D., Yeung, N., Heyes, C., and Frith, C. D. (2014). Supra-personal cognitive control and metacognition. Trends Cogn. Sci. 18, 186–193. doi:10.1016/j.tics.2014.01.006

Simon, D. A., and Bjork, R. A. (2001). Metacognition in motor learning. J. Exp.Psychol. Learn. Mem. Cogn. 27, 907–912. doi: 10.1037/0278-7393.27.4.907

Singer, R. N. (2000). Performance and human factors: considerations about cog-nition and attention for self-paced and externally-paced events. Ergonomics 43,1661–1680. doi: 10.1080/001401300750004078

Singer, R. N. (2002). Pre-performance state, routines, and automaticity: what does ittake to realize expertise in self-paced events? J. Sport Exerc. Psychol. 24, 359–375.

Stanley, J., and Krakauer, J. W. (2013). Motor skill depends on knowledge of facts.Front. Hum. Neurosci. 7:503. doi: 10.3389/fnhum.2013.00503

Stanovich, K. E. (2011). Rationality and the Reflective Mind. Oxford: OxfordUniversity Press.

Sternberg, R. (2001). “Metacognition, abilities, and developing expertise: whatmakes an expert student?” in Metacognition in Learning and Instruction, Vol.19, ed. H. J. Hartman (Dordrecht: Kluwer Academic Publishers), 247–260.

Sternberg, R. J., and Sternberg, K. (2012). Cognitive Psychology. Belmont, CA:Wadsworth.

Strack, F., and Bless, H. (1994). Memory for non-occurrences: metacog-nitive and presuppositional strategies. J. Mem. Lang. 33, 203–217. doi:10.1006/jmla.1994.1010

Tarricone, P. (2011). The Taxonomy of Metacognition. New York: Psychology Press.Toner, J. (2014). Knowledge of facts mediate “continuous improvement” in elite

sport: a comment on Stanley and Krakauer (2013). Front. Hum. Neurosci. 8:142.doi: 10.3389/fnhum.2014.00142

Toner, J., and Moran, A. (2011). The effects of conscious processing ongolf putting proficiency and kinematics. J. Sport. Sci. 29, 673–683. doi:10.1080/02640414.2011.553964

Toner, J., Moran, A., and Jackson, R. (2013). The effect of avoidant instructions ongolf putting proficiency and kinematics. Psychol. Sport Exerc. 14, 501–507. doi:10.1016/j.psychsport.2013.01.008

Van Overschelde, J. P. (2008). “Metacognition: knowing about knowing,” in Hand-book of Metamemory and Memory, eds J. Dunlosky and R. A. Bjork (New York:Psychology Press), 47–71.

Vickers, J. N. (2007). Perception, Cognition, and Decision Training: The Quiet Eye inAction. Champaign, IL: Human Kinetics.

Vine, S. J., Moore, L. J., and Wilson, M. R. (2011). Quiet eye training facili-tates competitive putting performance in elite golfers. Front. Psychol. 2:8. doi:10.3389/fpsyg.2011.00008

Vine, S. J., Moore, L. J., and Wilson, M. R. (2014). Quiet eye training: the acquisition,refinement and resilient performance of targeting skills. Eur. J. Sport Sci. 14(Suppl.1), S235–S242. doi: 10.1080/17461391.2012.683815

Vogt, S., Di Rienzo, F., Collet, C., Collins, A., and Guillot, A. (2013). Multiple rolesof motor imagery during action observation. Front. Hum. Neurosci. 7:807. doi:10.3389/fnhum.2013.00807

Voss, M. W., Kramer, A. F., Basak, C., Prakash, R. S., and Roberts, B. (2010).Are expert athletes“expert” in the cognitive laboratory? A meta-analytic reviewof basic attention and perception and sport expertise. Appl. Cogn. Psychol. 24,812–826. doi: 10.1002/acp.1588

Weil, L. G., Fleming, S. M., Dumontheil, I., Kilford, E. J., Weil, R. S., Rees, G.,et al. (2013). The development of metacognitive ability in adolescence. Conscious.Cogn. 22, 264–271. doi: 10.1016/j.concog.2013.01.004

Weinberg, R., Butt, J., Knight, B., Burke, K. L., and Jackson, A. (2003). The relation-ship between the use and effectiveness of imagery: an exploratory investigation.J. Appl. Sport Psychol. 15, 26–40. doi: 10.1080/10413200305398

Williams, A. M., and Ericsson, K. A. (2005). Perceptual-cognitive expertise in sport:some considerations when applying the expert performance approach. Hum.Mov. Sci. 24, 283–307. doi: 10.1016/j.humov.2005.06.002

Williams, A. M., and Ford, P. R. (2007). Expertise and sport performance. Int. Rev.Sport Exerc. Psychol. 1, 4–18. doi: 10.1080/17509840701836867

Wilson, T. D., and Gilbert, D. T. (2003). “Affective forecasting,” in Advances inExperimental Social Psychology, ed. M. Zanna (New York: Elsevier), 345–411.


Received: 19 June 2014; accepted: 24 September 2014; published online: 16 October2014.Citation: MacIntyre TE, Igou ER, Campbell MJ, Moran AP and Matthews J (2014)Metacognition and action: a new pathway to understanding social and cognitive aspectsof expertise in sport. Front. Psychol. 5:1155. doi: 10.3389/fpsyg.2014.01155This article was submitted to Cognition, a section of the journal Frontiers in Psychology.Copyright © 2014 MacIntyre, Igou, Campbell, Moran and Matthews. This is an open-access article distributed under the terms of the Creative Commons Attribution License(CC BY). The use, distribution or reproduction in other forums is permitted, providedthe original author(s) or licensor are credited and that the original publication in thisjournal is cited, in accordance with accepted academic practice. No use, distribution orreproduction is permitted which does not comply with these terms.







PERSPECTIVE ARTICLEpublished: 16 July 2014

doi: 10.3389/fpsyg.2014.00769

In praise of conscious awareness: a new framework for theinvestigation of “continuous improvement” in expertathletesJohn Toner 1* and Aidan Moran 2

1 School of Sport, Health and Exercise Sciences, University of Hull, Hull, UK2 School of Psychology, University College Dublin, Dublin, Ireland

Edited by:


Reviewed by:

John Sutton, Macquarie University,AustraliaDoris Jane McIlwain, MacquarieUniversity, Australia

*Correspondence:

John Toner, School of Sport, Healthand Exercise Sciences, University ofHull, Hull, HU6 7RX, UKe-mail: [email protected]

A key postulate of traditional theories of motor skill-learning (e.g., Fitts and Posner, 1967;Shiffrin and Schneider, 1977) is that expert performance is largely automatic in nature andtends to deteriorate when the performer “reinvests” in, or attempts to exert consciouscontrol over, proceduralized movements (Masters and Maxwell, 2008). This postulateis challenged, however, by recent empirical evidence (e.g., Nyberg, in press; Geeveset al., 2014) which shows that conscious cognitive activity plays a key role in facilitatingfurther improvement amongst expert sports performers and musicians – people whohave already achieved elite status (Toner and Moran, in press). This evidence suggeststhat expert performers in motor domains (e.g., sport, music) can strategically deployconscious attention to alternate between different modes of bodily awareness (reflectiveand pre-reflective) during performance. Extrapolating from this phenomenon, the currentpaper considers how a novel theoretical approach (adapted from Sutton et al., 2011) couldhelp researchers to elucidate some of the cognitive mechanisms mediating continuousimprovement amongst expert performers.

Keywords: expertise, continuous improvement, attention, embodiment, bodily awareness

Many traditional and contemporary theories of motor learningemphasize the apparently “spontaneous” nature of skilled perfor-mance. For example, Fitts and Posner (1967) and Shiffrin andSchneider (1977) argued that skilful action runs “automatically”or “procedurally.” Similarly, Dreyfus (2002) claimed that expertperformance proceeds “without calculating and comparing alter-natives . . . what must be done, simply is done” (p. 372). Commonto these accounts of expertise is the belief that any form of con-scious involvement during on-line skill execution is likely to provedeleterious to movement and performance proficiency (Beilocket al., 2002). Challenging these dominant perspectives on skill-learning, however, is an emerging body of theory (e.g., Breivik,2013; Toner and Moran, in press; Winter et al., 2014) and empiri-cal evidence (see Nyberg, in press) which suggests that“continuousimprovement” (i.e., the phenomenon whereby certain skilled per-formers appear to be capable of increasing their proficiency eventhough they are already experts) at the elite level of sport is heavilyreliant upon the performer’s ability to move efficiently betweenreflective and unreflective modes of bodily awareness. For exam-ple, experts are often required to pay conscious attention to, or“reinvest” in the training context, when their movements become“attenuated” (see Collins et al., 1999) because they believe that inorder to optimize their performance they must “experiment withand research their moving body” (Ravn and Christensen, in press).

Interestingly, evidence suggests that many skilled performersremain “somaesthetically” aware (i.e., focusing on the propriocep-tive “feel”, Shusterman, 2008) of their movement during on-lineskill execution (in the performance context; see Nyberg, in press)and can use global or holistic cue words to improve performance

proficiency under pressurized conditions (see Mullen and Hardy,2010). Therefore, instead of relying wholly on unthinking spon-taneity to guide their performance, elite athletes appear to alternatebetween different modes of cognitive processing, and also betweentypes of bodily self-awareness (i.e., reflective and pre-reflective)in practice and performance contexts. Here athletes might adoptwhat Colombetti (2011) refers to as a reflective awareness of theirbodily selves when they consider their intentions or actions andassess whether they are appropriate to a certain situation. By con-trast, pre-reflective awareness occurs when we are immersed in anactivity but our attention is not on our bodily selves. However,in the latter case, Colombetti (2011) argues that the body is notentirely invisible or absent from experience as it remains “as asource of feeling, affect, agency and expressivity” (p. 27).

In a recent paper (see Toner and Moran, in press), we arguedthat self-focused attention (including reflective bodily awareness)plays an important role in allowing skilled athletes to refine inef-ficient movements during deliberate practice. The present paperextends this argument by postulating that skilled performers arecapable of strategically allocating attention, and hence alternat-ing between reflective and pre-reflective modes of awareness, inorder to meet the requirements of dynamically unfolding and con-textually contingent performance environments. Furthermore, weargue that influential theoretical accounts (such as self-focus theo-ries; e.g., Beilock and Carr, 2001) used by researchers to identify thecognitive mechanisms underpinning performance at the expertlevel may be unable to adequately capture or explain the dynamical(i.e., that it may be freely allocated) nature of attentional process-ing amongst elite performers. Arising from these arguments, we













Toner and Moran In search of a new theoretical

propose that Sutton et al.’s (2011) “Applying intelligence to thereflexes” (AIR) approach, and a number of methodologies whichaim to uncover participants’ phenomenological descriptions oftraining and performance, may be better suited in achieving thislatter aim.

Over the last decade or so, experimental psychologists have usedstandardized laboratory tasks in order to identify the attentionalprocesses that govern skilled movement and performance in sport(e.g., Beilock and Carr, 2001; Jackson et al., 2006). For example,Beilock et al. (2002) found that when skilled golfers attended to aspecific aspect of their putting technique (e.g., the exact momentthat the clubhead finished its follow-through), their performancewas impaired relative to counterparts in a dual-task conditionwho putted while performing a secondary task (an auditory tone-monitoring activity). Further evidence for the detrimental effectsof skill-focused attention on skilled performance have been foundin a baseball batting task (Gray, 2004); in soccer dribbling (Fordet al., 2005); and in a golf putting task (Mullen and Hardy, 2000).According to self-focus theories of attention (including Beilockand Carr, 2001, “explicit monitoring hypothesis” and Masters,1992, “conscious processing hypothesis”) attending to the step-by-step component processes of a proceduralized skill resultsin its control structures being broken down into a sequence ofsmaller, separate, independent units. Ultimately, this creates theopportunity for error that was not present in the “chunked” con-trol structure (Beilock et al., 2002). Accordingly, performers havebeen advised to avoid focusing on their bodily movement and,instead, to “shift the focus to the external world, in particular onthe impact or effect of one’s behaviors” (Weiss and Reber, 2012,p. 174).

At first glance, it would seem that the case is quite clear – anyform of conscious processing that directs the performer’s atten-tion to their bodily movement is likely to disrupt fluent skillexecution. However, a closer look at the laboratory-based evi-dence base would suggest that its findings must be interpretedwith caution. To explain, in each of the aforementioned stud-ies athletes were asked to focus on a feature of their movementwhich they may never have previously focused on (and hencenever practiced doing so). Indeed, few studies have sought to cap-ture athletes’ attentional processes over time (e.g., between the“off-season” and competitive season) or across different situations(e.g., when recovering from injury). Given the dearth of stud-ies in psychology on the temporal and/or contextual dynamics ofattentional processes, we question the degree to which availableevidence supports the received wisdom that conscious attentionwill inevitably disrupt skilful performance. In fact, recent researchhas begun to cast doubt upon the validity of this latter assumption.For example, Nyberg (in press) found that elite freeskiers attendedto on-line skill execution in order to identify any features of theirmovement which might require alteration/adjustment in order tomaintain performance proficiency. In addition, a large volume ofevidence indicates that elite performers are capable of flexibly allo-cating their attention (i.e., moving from reflective to pre-reflectivemodes of bodily awareness) dependent upon the context-specificdemands confronting them during training and competitive per-formance (see Bernier et al., 2011) or the challenges (e.g., injuries,slumps) that they will inevitably face at some stage during their

careers (see Collins et al., 1999). In stark contrast to self-focus the-ories, these emergent findings suggest that skilled performance islikely to be impeded if the “proceduralization” of skills is excessive(see Ericsson, 2006) – because experts must be able to deliber-ately access and strategically re-route any semi-automated routinesin order to facilitate “continuous improvement” (Montero, 2010;Breivik, 2013).

Against this background, we argue that there is ample empir-ical evidence that “continuous improvement” at the elite level isheavily dependent upon the performer’s ability to effectively uti-lize reflective modes of bodily awareness. Skill-focused attention(including conscious bodily awareness) appears to be a key featureof skilled performance because athletes operating at this level aredriven by the desire to learn “new and better techniques” (Breivik,2007, p. 127). For example, despite having won eight medalsat the Beijing Olympics, Michael Phelps decided to change hisfreestyle technique in a bid to increase his sprinting speed (Ander-sson, 2009). Moreover, athletes will inevitably experience injury,fatigue, growth and aging which may disrupt habitual movement(see Bissell, 2013; Eden, 2013) and require them to correct, relearnand adjust their spontaneous performance (Shusterman, 2008).In fact, research on the topic of “skill recovery” or “skill refine-ment” shows that skilled athletes who are attempting to regainprior levels of performance often deliberately reinvest consciouscontrol to restore or refine habitual movements in sports suchas javelin throwing and swimming (Hanin et al., 2004). In thesestudies, researchers have helped athletes regain or refine disruptedmovement patterns by encouraging them to become more con-sciously aware of the somaesthetic differences between current(problematic) and desired actions.

Somaesthetic awareness appears to play an important role in“continuous improvement” by allowing performers to identifymovements that are causing them discomfort or outcomes whichare unusual or undesirable. Indeed, evidence suggests that theseforms of self- awareness are important mediators of “flow” oroptimal competitive performance in sport. On the basis of theirpioneering research on flow in sport, Jackson and Csikszentmi-halyi (1999) argued that “without self-awareness an athlete missesimportant cues that can lead to a positive change in performance”(p. 105). According to these authors, self-awareness simply meanspaying attention to cues provided by movements, and makingadjustments to your actions when something is amiss. Athletesmay use reflective bodily awareness to identify “attenuated” habitsin the performance context and subsequently adjust problematicmovements in the training context. However, evidence suggeststhat performers may also choose to adjust problematic move-ments during on-line execution during competition. To illustrate,Collins et al. (2001) found that elite weightlifters chose to con-sciously modify their movement during competition in order tomaintain movement proficiency. Similarly, Nyberg found that elitefreeskiers learn how to discern (i.e., through “focal awareness”)their rotational velocity to such an extent that they “know whetherthey will be able to perform the trick the way it was intendedwithout adjustments, or whether they will need to make adjust-ments during the flight phase” (p. 7). In this study, elite free skierswere video recorded during practice and subsequently interviewedusing a technique known as stimulated recall (SR – a method for





enhancing reflection by recalling situations through audiotapesor video recordings). Nyberg suggests that these performers relyon their focal awareness (which is conscious and might includeknowledge of their velocity and how they need to modify it) andtheir subsidiary awareness which is “less conscious” and includesknowledge of the “particulars” such as the friction of the snow andtheir feelings of previous jumps. These elite performers were foundto navigate their focal awareness by rapidly shifting its target evenin the midst of the activity itself. For one performer, this meantthat he was focally and embodiedly aware of his rotational velocitywhile in the air but could quickly change his awareness to take intoaccount environmental conditions such as his position in relationto the targeted landing area. Clearly, these findings suggest thatsome performers seek to counteract automaticity by ensuring thatcertain features of performance are subject to strategic control.

Although some performers may choose to reinvest consciousattention by adjusting movement in the performance context, oth-ers may choose to use cue words as “instructional nudges” (i.e.,explicit verbal phrases or maxims: see Sudnow, 2001; Sutton, 2007)in order to“re-route”embodied routines. According to Sutton et al.(2011) cues may allow the performer to build and access “flexiblelinks between knowing and doing” (p. 95). Cue words appearto represent forms of thinking and remembering which can, insome circumstances, allow performers to animate the kinestheticmechanisms of skilled performance. To illustrate, Jenkins (2007)interviewed 113 European tour golfers and found that 70% ofthese performers used at least one “swing thought” (i.e., a form ofcue word) during on-line skill execution. Clearly, certain forms ofmindedness or conscious processing are a common feature of elitecompetitive performance.

The preceding evidence would suggest that a better understand-ing of the cognitive processes mediating continuous improvementat the elite level of skilled performance can be achieved onlyby adopting a theoretical framework which can account for thedynamic nature of attentional processing. Therefore, insteadof explanations (e.g., self-focus theories) that emphasize theproceduralized nature of skilled performance, we may require the-oretical accounts which focus on the interchanging phases or stagesof learning (Shusterman, 2008, 2009) that appear to characterizetraining and performance at the elite level of sport. Consequently,we propose that Sutton et al.’s (2011) AIR approach may helpresearchers to explain how performers can alternate between dif-ferent modes of processing. Briefly, Sutton et al.’s (2011) modelis cyclical in the sense that the maintenance and enhancement ofperformance efficiency requires the“rapid switching of modes andstyles” (2011, p. 93) within the training and performance context.This framework proposes that expert skill relies on a mindednessthat“facilitates the dynamic flexibility of attention, allowing it to beallocated freely and in a way that best meets contingent contextualdemands” (Geeves et al., 2014, p. 676). Accordingly, Geeves et al.(2014) argue that expert performers may determine the amountof attention they need to pay to certain processes in the practicecontext (depending on their current level of performance) andduring on-line performance (according to the situational demandspresented to them).

Additionally, we propose that the AIR model may helpresearchers to interpret the accumulating body of empirical

evidence which suggests that skilled performers seek to avoid auto-maticity by ensuring that performance remains open to strategiccontrol in both the practice and performance context. Suttonet al. (2011) argued that there are a number of different ways inwhich embodied coping is minded or mindful (varying acrossindividuals, task domains, and cultures) and recommend thatwe search for forms of theorizing that highlight these differ-ences by exploring what actually happens to performers as they“direct attention to kinesthetic cues in increasingly skilful ways”(p. 96). Given the preceding evidence documenting the mind-ful nature of expertise, it would also seem important to identifymethodological approaches that will help researchers capturethe attentional switching mechanisms that appear to underpin“continuous improvement” at the elite level of sport. Specifi-cally, in order to understand embodied perspectives of experienceresearchers may wish to adopt methodological approaches that are“truly grounded in the carnal realities of the lived sporting bod-ies” (Hockey and Allen-Collinson, 2007, p. 116). Some researchershave recently taken up this challenge by drawing on participants’phenomenological insights to better understand how embodiedexpertise is shaped by training and performance. For example,Ravn and Christensen (in press) utilized a phenomenology-relatedanalysis of qualitative data (including participant observations andinterviews) with an elite golfer to explore how the “describedexperience comes into being rather than what this experiencemeans to the subject” (p. 5). The principal author observedLine (an elite golfer), between 5 and 8 h each day, over 5 daysof training. The researchers drew on notes taken during theseobservations to invite Line to describe her practice sessions andexperiences in detail. The researchers subsequently analyzed thedata by looking for “petite generalizations” (i.e., generalizationsthat regularly occurred in the case study) relating to how Lineused her awareness of bodily sensations during training. Over-all, the findings suggested that training at the elite level is notjust about handling the physicality of the body but also aboutlistening to it and regulating how it should “feel” in order to per-form optimally (see Nyberg, in press, for a similar methodologicalapproach).

In another study, Bernier et al. (2011) used a naturalistic inves-tigation to explore the attentional foci adopted by elite golfers intraining and performance contexts. The initial phase of the studyinvolved filming participants (for 60 min) in a training sessionduring the winter (non-competitive) season. The second phasetook place during the competitive season and involved filmingparticipants over the course of a complete round (i.e., 18 holesof play) in a professional competition. Self-confrontation inter-views, based on video footage, were conducted within 2 h of thecompletion of the training session and competitive performance.The interviews sought to encourage SR by asking each partici-pant to view the videotape with the researcher and to recall anddescribe what thoughts he was processing during the training ses-sion or competition. Initially, the participants and the researcherwatched the first situations on the video recording (i.e., the firsttraining exercise and the first three holes of the competition).Having discussed the attentional foci adopted during these situa-tions, the participant was asked to indicate specific circumstancesthat he considered relevant to analyse. These situations included





specific exercises during the training sessions or great/poorly exe-cuted strokes during competition. Each sequence involved anaction (e.g., a practice drill or a shot), the preparatory phase(e.g., the pre-shot routine), and the step following this action(e.g., walking to the next shot). Participants were urged to expresstheir thoughts during each sequence, rather than being asked toexplain “their solution for the task or to provide a summary ofthe general strategy adopted” (Bernier et al., 2011, p. 331). Induc-tive content analysis revealed that these elite golfers adapted theirattentional focus depending on the context. That is, golfers werefound to flexibly adjust their attentional focus (moving back-and-forth between internal and external foci) across the preparatory,execution and evaluative stages of training and competitiveperformance.

In summary, a significant volume of evidence shows thatskilled performers’ foci of attention may change dramatically overthe course of a competitive season (e.g., to deal with “attenu-ated” movement patterns) or during a competitive event (i.e.,between preparation, execution, and evaluation). Unfortunately,as most studies of attentional processes in psychology are lim-ited to static snapshots of the phenomena of interest, they shedlittle light on the dynamical nature of attention. Against thisbackground, the current paper has argued how Sutton et al.’s(2011) AIR approach may provide researchers with a more detailedunderstanding of the embodied nature of skilful action andperformance – thereby helping to explain how athletes strate-gically allocate attentional resources in seeking to maintain andenhance performance proficiency. In order to shed light onthe complex and dynamic attentional mechanisms that mediate“continuous improvement” in expert performers researchers mayneed to use a variety of methods including both standardizedlaboratory techniques (e.g., occlusion paradigms, eye-tracking)and phenomenological and naturalistic investigations (e.g., SR).Together, these approaches may help researchers understand howand why performers alternate between reflective and pre-reflectivemodes of bodily awareness across training and performancecontexts.

REFERENCESAndersson, A. (2009). Michael Phelps Changes Technique in Bid for Perfect Stroke.

Available at: http://www.telegraph.co.uk/sport/olympics/swimming/5205810/Michael-Phelps-changes-technique-in-bid-for-perfect-stroke.html

Beilock, S. L., and Carr, T. H. (2001). On the fragility of skilled performance:what governs choking under pressure? J. Exp. Psychol. Gen. 130, 701–725. doi:10.1037/0096-3445.130.4.701

Beilock, S. L., Carr, T. H., MacMahon, C., and Starkes, J. L. (2002).When paying attention becomes counterproductive: impact of divided ver-sus skill-focused attention on novice and experienced performance of sen-sorimotor skills. J. Exp. Psychol. Appl. 8, 6–16. doi: 10.1037/1076-898X.8.1.6

Bernier, M., Codron, R., Thienot, E., and Fournier, J. F. (2011). The attentional focusof expert golfers in training and competition: a naturalistic investigation. J. Appl.Sport Psychol. 23, 326–341. doi: 10.1080/10413200.2011.561518

Bissell, D. (2013). Habit displaced: the disruption of skilful performance. Geogr. Res.51, 120–129. doi: 10.1111/j.1745-5871.2012.00765.x

Breivik, G. (2007). Skillful coping in everyday life and in sport: a critical exami-nation of the views of Heidegger and Dreyfus. J. Philos. Sport 34, 116–134. doi:10.1080/00948705.2007.9714716

Breivik, G. (2013). Zombie-like or superconscious? a phenomenological and con-ceptual analysis of consciousness in elite sport. J. Philos. Sport 40, 85–106. doi:10.1080/00948705.2012.725890

Colombetti, G. (2011). Varieties of pre-reflective awareness: foregroundand background bodily feelings in emotion. Inquiry 54, 293–313. doi:10.1080/0020174X.2011.575003

Collins, D., Jones, B., Fairweather, M., Doolan, S., and Priestley, N. (2001). Exam-ining anxiety associated changes in movement patterns. Int. J. Sport Psychol. 32,223–242.

Collins, D., Morriss, C., and Trower, J. (1999). Getting it back: a case study of skillrecovery in an elite athlete. Sport Psychol. 13, 288–298.

Dreyfus, H. (2002). Intelligence without representation: Merleau Ponty’scritique of representation. Phenomenol. Cogn. Sci. 1, 367–383. doi:10.1023/A:1021351606209

Eden, S. (2013). Stroke of Madness: How has Tiger Woods Managed to Overhaul hisSwing Three Times? Available at: http://espn.go.com/golf/story//id/8865487/tiger-woods-reinvents-golf-swing-third-time-career-espn-magazine

Ericsson, K. A. (2006). “The influence of experience and deliberate practice onthe development of superior expert performance,” in Cambridge Handbook ofExpertise and Expert Performance, eds K. A. Ericsson, N. Charness, P. Feltovich,and R. R. Hoffman (Cambridge: Cambridge University Press), 685–706.

Fitts, P. M., and Posner, M. I. (1967). Human Performance. California: Brooks/ColePublishing Company.

Ford, P., Hodges, N. J., and Williams, A. M. (2005). Online attentional-focus manip-ulations in a soccer dribbling task: implications for the proceduaralization ofmotor skills. J. Motor Behav. 37, 386–394. doi: 10.3200/JMBR.37.5.386-394

Geeves, A., McIlwain, D. J. F., Sutton, J., and Christensen, W. (2014). To think or notto think: the apparent paradox of expert skill in music performance. Educ. Philos.Theory 46, 674–691. doi: 10.1080/00131857.2013.779214

Gray, R. (2004). Attending to the execution of a complex sensorimotor skill: Exper-tise differences, choking, and slumps. J. Exp. Psychol. Appl. 10, 42–54. doi:10.1037/1076-898X.10.1.42

Hanin, Y., Malvela, M., and Hanina, M. (2004). Rapid correction of start techniquein an Olympic-level swimmer: a case study using old way/new way. J. Swim. Res.16, 11–17.

Hockey, J., and Allen-Collinson, J. (2007). Grasping the phenomenology of sportingbodies. Int. Rev. Soc. Sport 42, 115–131. doi: 10.1177/1012690207084747

Jackson, R. C., Ashford, K. J., and Norsworthy, G. (2006). Attentional focus, dis-positional reinvestment, and skilled motor performance under pressure. J. SportExerc. Psychol. 28, 49–68.

Jackson, S. A., and Csikszentmihalyi, M. (1999). Flow in Sports: The Keys to OptimalExperiences and Performances. Champaign, IL: Human Kinetics.

Jenkins, S. (2007). The use of swing keys by elite tournament professional golfers.Annu. Rev. Golf Coach. 1, 199–217. doi: 10.1260/174795407789705460

Masters, R. S. W. (1992). Knowledge, knerves and know-how: the role of explicitversus implicit knowledge in the breakdown of a complex motor skill underpressure. Br. J. Psychol. 83, 343–358. doi: 10.1111/j.2044-8295.1992.tb02446.x

Masters, R. S. W., and Maxwell, J. (2008). The theory of reinvestment. Int. Rev. SportExerc. Psychol. 1, 160–183. doi: 10.1080/17509840802287218

Montero, B. (2010). Does bodily awareness interfere with highly skilledmovement? inquiry: an interdisciplinary. J. Philos. 53, 105–122. doi:10.1080/00201741003612138

Mullen, R., and Hardy, L. (2000). State anxiety and motor performance: test-ing the conscious processing hypothesis. J. Sports Sci. 18, 785–799. doi:10.1080/026404100419847

Mullen, R., and Hardy, L. (2010). Conscious processing and the process goalparadox. J. Sport Exerc. Psychol. 32, 275–297.

Nyberg, G. (in press). Developing a “somatic velocimeter” – the practical knowledgeof freeskiers. Qual. Res. Sport Exerc. Health. doi: 10.1080/2159676X.2013.857709

Ravn, S., and Christensen, M. K. (in press). Listening to the body? how phenomeno-logical insights can be used to explore a golfer’s experiences of the physicality ofher body. Qual. Res. Sport Exerc. Health doi: 10.1080/2159676X.2013.809378

Shiffrin, R. M., and Schneider, W. (1977). Controlled and automatic human infor-mation processing: II. perceptual learning, automatic attending, and a generaltheory. Psychol. Rev. 84, 127–190. doi: 10.1037/0033-295X.84.2.127

Shusterman, R. (2008). Body Consciousness: A Philosophy of Mindful-ness and Somaesthetics. Cambridge: Cambridge University Press. doi:10.1017/CBO9780511802829

Shusterman, R. (2009). Body consciousness and performance: somaestheticseast and west. J. Aesthet. Art Criticism 67, 133–145. doi: 10.1111/j.1540-6245.2009.01343.x


http://www.telegraph.co.uk/sport/olympics/swimming/5205810/Michael-Phelps-changes-technique-in-bid-for-perfect-stroke.html

http://www.telegraph.co.uk/sport/olympics/swimming/5205810/Michael-Phelps-changes-technique-in-bid-for-perfect-stroke.html

http://espn.go.com/golf/story//id/8865487/tiger-woods-reinvents-golf-swing-third-time-career-espn-magazine

http://espn.go.com/golf/story//id/8865487/tiger-woods-reinvents-golf-swing-third-time-career-espn-magazine




Sudnow, D. (2001). Ways of the Hand: A Rewritten Account. Cambridge, MA: MITpress.

Sutton, J. (2007). Batting, habit and memory: the embodied mind and the natureof skill. Sport Soc. 10, 763–786. doi: 10.1080/17430430701442462

Sutton, J., Mcilwain, D., Christensen, W., and Geeves, A. (2011). Applying intelli-gence to the reflexes: embodied skills and habits between dreyfus and descartes.J. Br. Soc. Phenomenol. 42, 78–103.

Toner, J., and Moran, A. (in press). Toward an explanation ofcontinuous improve-ment in expert athletes: the role of consciousness in deliberate practice. Int. J.Sport Psychol.

Weiss, S. M., and Reber, A. S. (2012). Curing the dreaded “Steve Blass Disease”. J.Sport Psychol. Action 3, 171–181. doi: 10.1080/21520704.2012.682702

Winter, S., MacPherson, A. C., and Collins, D. (2014). “To think, or not tothink, that is the question”. Sport Exerc. Perform. Psychol. 3, 102–115. doi:10.1037/spy0000007


Received: 28 April 2014; accepted: 01 July 2014; published online: 16 July 2014.Citation: Toner J and Moran A (2014) In praise of conscious awareness: a new frame-work for the investigation of “continuous improvement” in expert athletes. Front.Psychol. 5:769. doi: 10.3389/fpsyg.2014.00769This article was submitted to Cognition, a section of the journal Frontiers in Psychology.Copyright © 2014 Toner and Moran. This is an open-access article distributed underthe terms of the Creative Commons Attribution License (CC BY). The use, distributionor reproduction in other forums is permitted, provided the original author(s) or licensorare credited and that the original publication in this journal is cited, in accordance withaccepted academic practice. No use, distribution or reproduction is permitted whichdoes not comply with these terms.






PERSPECTIVE ARTICLEpublished: 08 July 2014

doi: 10.3389/fpsyg.2014.00709

An adaptive toolbox approach to the route to expertise insportRita F. de Oliveira1*, Babett H. Lobinger 2 and Markus Raab1,2

1 Department of Applied Sciences, London South Bank University, London, UK2 Institute of Psychology, German Sport University, Cologne, Germany

Edited by:

Merim Bilalic, University ofKlagenfurt, Austria

Reviewed by:

Bartosz Gula, University of Klagenfurt,AustriaMerim Bilalic, University ofKlagenfurt, Austria

*Correspondence:

Rita F. de Oliveira, Department ofApplied Sciences, London South BankUniversity, 103 Borough Road,London, SE10AA, UKe-mail: [email protected]

Expertise is characterized by fast decision-making which is highly adaptive to newsituations. Here we propose that athletes use a toolbox of heuristics which they developon their route to expertise. The development of heuristics occurs within the context ofthe athletes’ natural abilities, past experiences, developed skills, and situational context,but does not pertain to any of these factors separately. This is a novel approach becauseit integrates separate factors into a comprehensive heuristic description. The novelty ofthis approach lies within the integration of separate factors determining expertise into acomprehensive heuristic description. It is our contention that talent identification methodsand talent development models should therefore be geared toward the assessment anddevelopment of specific heuristics. Specifically, in addition to identifying and developingseparate natural abilities and skills as per usual, heuristics should be identified anddeveloped. The application of heuristics to talent and expertise models can bring the fieldone step away from dichotomized models of nature and nurture toward a comprehensiveapproach to the route to expertise.

Keywords: heuristics, expertise, sport, talent development, talent identification, cues

INTRODUCTIONMost current theories of expertise are based on the principle thatknowledge underlies performance (Ericsson et al., 2006). Specif-ically, Ericsson et al. (2006) suggest that the specific knowledgeof experts arises through about 10.000 h of deliberate practice.More research studies have also shown that previous knowl-edge guides attention for accurate performance (e.g., Bilalic et al.,2010). Drawing on the importance of knowledge for performance,the heuristic approach focuses on how that knowledge can beeffectively searched and a how solution implemented.

Heuristics are rules of thumb that allow fast and frugal decision-making. The concept of heuristics was introduced by Simon (1982)to explain how humans decide when they have limited resources.He proposed that behavior could only be understood throughanalyzing both the person and the environment where the behav-ior took place. The subsequent work of Gigerenzer et al. (1999)identified and tested specific heuristics in a number of differentenvironments. For instance, they found the recognition heuristicwhereby people choose the option they recognize over the optionthey do not recognize (Gigerenzer and Goldstein, 1999). Recently,the simple heuristics research program has been used in the contextof sport. Raab (2012) showed that athletes use simple heuristicsboth to make decisions and to implement them in the sports envi-ronment. What is still lacking, however, is an understanding ofhow simple heuristics develop in the route to expertise.

HEURISTICS AS CHARACTERISTICS OF EXPERTISE AND TALENTDEVELOPMENTThe topic of expertise has been gaining increased prominencein science and the media because researchers, practitioners, and

laypeople wish to replicate the route to success in the most effi-cient way. In sport, and especially in team sports, experts arethose with repeated top-level performance who can most effi-ciently resolve the situation put before them. The route to that levelof expertise will no doubt have involved uncountable attemptssome successful but many unsuccessful. Here we will argue thatthe developmental process is inherently non-optimal and non-linear, but that this is indispensable to acquire the highest levelsof expertise. To say that athletes show optimal adaptation to var-ious situations related to their sport does not equate to sayingthat their actions or behaviors are optimal, but rather that theseare cost-effective actions or behaviors. This is especially the casewhere performance involves fast decision-making. The best deci-sion is not the optimal decision per se, but the one that can solvethe current situation well enough and fast enough (Simon, 1982;Raab et al., 2009; Todd et al., 2012). It is important to qualifywhat is meant by optimal performance so that efforts to iden-tify and develop talented athletes are geared toward functional(not optimal) decision-making. The route to expertise starts bydemonstrating a talent. Talent identification is the process of rec-ognizing the potential of an athlete to excel in a particular sport.Talent development, on the other hand, is the process by whichan athlete can realize that potential, which includes benefitingfrom the most appropriate learning and training environments(Vaeyens et al., 2008).

CLASSICAL MODELS IN TALENT DEVELOPMENT: NATURAL ABILITIESAND NURTURENatural abilities together with environmental and intrapersonalcatalysts are, according to Gulbin et al. (2010), the non-random


http://www.frontiersin.org/Psyrrchology/












de Oliveira et al. Adaptive toolbox approach to expertise in sport

factors contributing to the developmental process which leadsto specific competencies. The distinction between natural abil-ities and catalysts reflects current views on talent identificationand development which show a facet of the nature vs. nur-ture debate (e.g., Baker et al., 2003; Epstein, 2013). The focusof this debate is on general natural abilities and environmen-tal factors that lead to general competencies (cf. categorizationsby Gagné, 1999). For instance perceptiveness and coordinationare considered general natural abilities and thought to influencetalent development. Athletes are often identified on general per-ceptual or motor skills, and their development often focuses onthese general abilities. Paradoxically, researchers and practition-ers agree that expert athletes specialize in their sport and thata general skill is not sufficient for expert performance. One ofthe marks of expertise is the use of unique solutions to solvesituations in the playfield and in other situations related withthe sport. In other words, expert athletes are characterized bytheir optimal adaptation to all things related with their sport,including effective decision-making in situ rather than by generalabilities.

THE HEURISTICS APPROACH TO TALENT DEVELOPMENTA useful framework to understand unique adaptations to newsituational contexts is the heuristic approach. Athletes use heuris-tics or rules of thumb which are specific to the type of situationand can be used rapidly without much cost. The developmentof heuristics occurs within the context of the athletes’ naturalabilities, but also their past experiences, developed skills, andsituational contexts. It does not pertain to any of these factorsseparately, instead, heuristics pertain to the repertoire of theathlete and it is our contention that talent identification meth-ods and talent development models should be geared toward theassessment and development of heuristics. This can be done byimproving the efficiency in the use of an existing heuristic (e g.,calibrating the heuristic to more valid cues), or learning whichheuristic fits best with a specific environment (e g., Gigeren-zer and Gaissmaier, 2011). The heuristics repertoire consists ofpsychological, neurophysiological, and perceptual-motor adap-tations (Raab et al., 2009; Raab, 2012; Todd et al., 2012). Eachheuristic is used for specific situations in much the same wayas a hammer is used for nailing pictures but not for cuttingbranches. By definition a heuristic is composed of at least threebuilding blocks. They are search rules, search-stopping rules, anddecision rules. Within sports a fourth building block has been pro-posed which deals with the execution rules. We will expand onthese.

Search rules include two kinds of search: search for informa-tion cues and search for alternatives. In most ball games thealternatives are fixed (i.e., players can only either pass, drib-ble, throw, etc.) so the athlete searches for information cuesto decide on which of these actions to use. While novice ath-letes may search for information cues randomly, expert athletescan directly use the information cues with the highest validity(de Oliveira et al., 2009; Esteves et al., 2011). When the alterna-tive actions are not specified, then search rules for the actionitself must be generated. In these cases the task characteristicsspecify whether it is most advantageous to broaden or limit the

search. For instance in chess it is advantageous to broaden thesearch for options (Bilalic et al., 2009), whereas in more time-pressured sports it may be best to narrow the search for options(Raab and Johnson, 2007).

Search-stopping rules are the rules by which one stops search-ing for information cues or alternative actions. Classical modelspresumed that there was a way to compute the optimal stop-ping point where the costs of further search would exceedits benefits. However, to say that athletes show optimal useof heuristics does not equate to saying that their actions orbehaviors are optimal, but rather that these are cost-effectiveactions or behaviors. This means that an expert athlete knowswhen the search for information cues must stop and willuse whatever information was gathered to make the decisionin due time. This also means that novice athletes must beplaced in situations that potentiate their search for the mostvalid information cues, and must also be placed in situationswhere decisions must be made based on low-quality informationcues.

Decision rules describe how a decision is made after search hasbeen stopped. Decision rules define how the available informa-tion is used to make a decision. Psychology has a tradition ofassuming that intelligent behavior implies weighting and com-bining information cues (e.g., multiple linear regression models),but the research on fast and frugal heuristics has shown that fre-quently less is more. For instance, the recognition heuristic is adecision rule whereby the option chosen is simply based on onevalid cue that point to one option and not to an alternative option(Gigerenzer and Goldstein, 1999). Again, expert athletes have theheuristic repertoire to make the best decisions the fastest, whereasnovice athletes must be placed in situations that build up theirrepertoire.

Execution rules address questions like what action to carryout and how to execute it as already described for decisionprocesses (Raab et al., 2005). Those rules are based on indi-vidual experience (Raab and Johnson, 2007). Athletes shouldbe exposed to situations that force them to decide betweenoptions to learn execution criteria and build heuristics for varioussituations.

These rules are the building blocks of heuristics and they canhelp explain how athletes develop their expertise in terms ofdecision-making and problem-solving which are key competencesin expert performance.

HEURISTICS IN THE ROUTE TO SPORT EXPERTISE: BUILDING ANADAPTIVE TOOLBOXHeuristics are domain-specific and can be used to formallydescribe the link between natural and nurtured characteristics.In fact heuristics are neutral in the nature vs. nurture debatebecause they can be learned but they can also be available at birth(cf., Baker et al., 2003, 2012). As an illustration, very small chil-dren will naturally cluster around a ball. The building blocks ofthe heuristic used might be: search for ball, stop searching whenthe ball is found, decide to move closer to the ball. This behav-ior continues until they learn that exploring the space increasesthe chance of receiving the ball. Here they may change the rulesof the existing heuristic into: search for space in relation to the





ball, stop when space is found, and decide to move the space.Again, this behavior will continue until they learn that not onlyspace but also the defensive players are important in receivingthe ball. Here they may again change the rules of the existingheuristic into: search for the space in relation to the ball and thedefensive player, stop when you found it decide to move to thespace.

Heuristics are also specific to the sportive situations that theathlete encounters. Therefore, nature and nurture, which arenormally described separately in models of talent identificationand development, can instead be described as heuristics. Thiswould allow future research on expertise to conceptually integratephylogenetic (natural abilities), ontogenetic (development) andsituational factors. As an illustration, take the phenomenon ofless-is-more for situations where there is an abundance of cues.We explore how athletes with different natural abilities (John andMary) may become experts in their sport. John may have a natu-ral ability to focus on a small number of cues and use those cuesto maximal advantage. He will be identified as talented becauseof his consistent results in particular situations (rather than hiscreativity). During development, John may specialize in the useof those cues and become an expert in using them and hencebuild a narrow repertoire of expertise. Provided these are themost valid cues for the sports situation, John will also be anexpert in his sport. If the sport offers a lot more variety, how-ever, John will need to benefit from a varied training programthat forces him to explore and use other cues for other situa-tions. Mary, on the other hand, may have a natural ability to focuson a large number of cues and will therefore use various com-binations of cues. She might be identified as talented because ofher creative solutions (rather than her consistent results). Dur-ing development, Mary may learn which cues are most valid towhich situation and hence build a broad repertoire of expertise.Provided a number of cue combinations is required for the sportssituation, Mary will also be an expert in her sport. If the sportoffers little variety, however, Mary will need to benefit from a spe-cialized training program that forces her to use the most validcues.

CONCLUSIONThe heuristics approach to expertise is useful because it takes intoaccount the natural abilities and development of the athlete, aswell as the situations posed by the sport and the training envi-ronment. It partly addresses the eternal nature vs. nurture debateand can provide suggestions for training programs which aim atdeveloping the individual natural abilities of athletes by providingadequate sports situations. This is currently being developed in thearea of decision-making (Marasso et al., accepted). The practicalapplication of heuristics to talent identification and talent develop-ment models will bring the field one step away from dichotomizedmodels and toward a true comprehensive approach to the routeto expertise. Future research can use the heuristics approach toinvestigate how the route to expertise sometimes deviates from themainstream to create novel solutions. For instance, new techniqueslike the Fosbury flop in track-and-field and the Tsukahara’s vaultin gymnastics (Bar-Eli et al., 2008) highlight alternatives found byexpert athletes who did not fit a standard model of talent.

ACKNOWLEDGMENTSThe three authors made substantial contributions to: the con-ception of the manuscript; its critical revision for importantintellectual content; final approval of the version submitted forpublication. The three authors agree to be accountable for allaspects of the work in ensuring that questions related to theaccuracy or integrity of any part of the work are appropriatelyinvestigated and resolved.

REFERENCESBaker, J., Cobley, S., and Schorer, J. (2012). Talent Identification and Development in

Sport: International Perspectives. New York, NY: Routledge.Baker, J., Horton, S., Robertson-Wilson, J., and Wall, M. (2003). Nurturing sport

expertise: factors influencing the development of elite athletes. J. Sports Sci. Med.2, 1–9.

Bar-Eli, M., Lowengart, O., Tsukahara, M., and Fosbury, R. D. (2008).Tsukahara’s vault and fosbury’s flop: a comparative analysis of two greatinventions. Int. J. Innov. Manag. 12, 21–39. doi: 10.1142/S136391960800190X

Bilalic, M., McLeod, P., and Gobet, F. (2009). Specialization effect and its influenceon memory and problem solving in expert chess players. Cogn. Sci. 33, 1117–1143.doi: 10.1111/j.1551-6709.2009.01030.x

Bilalic, M., Langner, R., Erb, M., and Grodd, W. (2010). Mechanisms and neural basisof object and pattern recognition – A study with chess experts. J. Exp. Psychol.Gen. 134, 728–742. doi: 10.1037/a0020756

de Oliveira, R. F., Oudejans, R. D., and Beek, P. J. (2009). Experts appear to use angleof elevation information in basketball shooting. J. Exp. Psychol. Hum. Percept.Perform. 35, 750–761. doi: 10.1037/a0013709

Epstein, D. (2013). The Sports Gene. London: Yellow Jersey.Ericsson, K. A., Charness, N., Feltovich, P. J., and Hoffman, R. R. (2006).

The Cambridge Handbook of Expertise and Expert Performance. New York, NY:Cambridge University Press. doi: 10.1017/CBO9780511816796

Esteves, P., de Oliveira, R. F., and Araújo, D. (2011). Posture-related affor-dances guide attack in basketball. Psychol. Sport Exerc. 12, 639–644. doi:10.1016/j.psychsport.2011.06.007

Gagné, F. (1999). My convictions about the nature of abilities, gifts, and talents. J.Educ. Gift. 22, 109–136.

Gigerenzer, G., and Gaissmaier, W. (2011). Heuristic decision making. Annu. Rev.Psychol. 62, 451–482. doi: 10.1146/annurev-psych-120709-145346

Gigerenzer, G., and Goldstein, D. G. (1999). “Betting on one good reason: the takethe best heuristic,” in Simple Heuristics that Make us Smart, eds G. Gigerenzer,P. M. Todd, and The ABC Research Group (New York, NY: Oxford University),75–95.

Gigerenzer, G., Todd, P. M., and the ABC Research Group. (1999). Simple Heuristicsthat Make us Smart. New York, NY: Oxford University.

Gulbin, J. P., Oldenziel, K. E., Weissensteiner, J. R., and Gagné, F. (2010). A lookthrough the rear view mirror: developmental experiences and insights of highperformance athletes. Talent Dev. Excell. 2, 149–164.

Raab, M. (2012). Simple heuristics in sports. Int. Rev. Sport Exerc. Psychol. 5,104–120. doi: 10.1080/1750984X.2012.654810

Raab, M., and Johnson, J. (2007). Expertise-based differences in search andoption-generation strategies. Exp. Psychol. Appl. 13, 158–170. doi: 10.1037/1076-898X.13.3.158

Raab, M., de Oliveira, R. F., and Heinen, T. (2009). How do people perceive andgenerate options? Prog. Brain Res. 174, 49–59. doi: 10.1016/S0079-6123(09)01305-3

Raab, M., Masters, R. S. W., and Maxwell, J. P. (2005). Improving the “how” and“what” decisions of elite table tennis players. Hum. Mov. Sci. 24, 326–344. doi:10.1016/j.humov.2005.06.004

Simon, H. A. (1982). Models of Bounded Rationality. Cambridge, MA: MIT Press.Todd, P. M., Gigerenzer, G., and ABC Research Group. (2012). Ecological Ratio-

nality: Intelligence in the World. New York, NY: Oxford University Press. doi:10.1093/acprof:oso/9780195315448.001.0001

Vaeyens, R., Lenoir, M., Williams, A. M., and Philippaerts, R. M. (2008). Tal-ent identification and development programmes in sport: current models andfuture directions. Sports Med. 38, 703–7714. doi: 10.2165/00007256-200838090-00001






Received: 01 April 2014; accepted: 19 June 2014; published online: 08 July 2014.Citation: de Oliveira RF, Lobinger BH and Raab M (2014) An adaptive tool-box approach to the route to expertise in sport. Front. Psychol. 5:709. doi:10.3389/fpsyg.2014.00709

This article was submitted to Cognition, a section of the journal Frontiers in Psychology.Copyright © 2014 de Oliveira, Lobinger and Raab. This is an open-access arti-cle distributed under the terms of the Creative Commons Attribution License(CC BY). The use, distribution or reproduction in other forums is permitted,provided the original author(s) or licensor are credited and that the original pub-lication in this journal is cited, in accordance with accepted academic practice. Nouse, distribution or reproduction is permitted which does not comply with theseterms.








Psychological perspectives on expertise

Documents

Transcript of Psychological perspectives on expertise