87 (2006) 289–296www.elsevier.com/locate/schres

Schizophrenia Research

Mixed handedness is associated with the Disorganizationdimension of schizotypy in a young male population

Nicholas C. Stefanis a,b,c,⁎, Silia Vitoratou a,d, Nikos Smyrnis a,b,Theodoros Constantinidis a, Ioannis Evdokimidis a, Ioannis Hatzimanolis b,

Ioannis Ntzoufras d, Costas N. Stefanis a

a University Mental Health Research Institute (UMHRI) 2 Soranou tou Efessiou str. P.O. BOX 66517, 156 01 Papagou, Athens, Greeceb Department of Psychiatry, National and Kapodistrian University of Athens, Greece, Eginition Hospital, 74 Vas. Sofias Ave., 11528 Athens, Greece

c King's College London, Institute of Psychiatry, Department of Psychological Medicine, London, UKd Department of Statistics, Athens University of Economics and Business, 76 Patision Str, 10434, Athens, Greece

Received 29 January 2006; received in revised form 19 April 2006; accepted 25 April 2006Available online 22 June 2006

Abstract

Within the ASPIS (Athens Study of Psychosis Proneness and Incidence of Schizophrenia) we sought out to examine inaccordance with previous reports if a deviation from dextrality is associated with an augmented endorsement of self ratedschizotypal personality traits in a large population of 1129 young male army recruits. Schizotypal traits were assessed using theSchizotypal Personality Questionnaire and hand preference membership was determined by applying stringent criteria derived fromthe Annett Handedness Questionnaire and the Porac–Coren questionnaire of lateral preferences. By adopting three differentdefinitions of hand preference membership, we confirmed an association between mixed handedness and increased schizotypalpersonality traits, and in particular with Disorganization schizotypy that encompasses aspects of self perceived difficulties in verbalcommunication. Non-verbal cognitive ability, as indexed by measurement of non-verbal IQ, sustained attention and workingmemory was not associated with hand preference. We argue that a deviation from normal cerebral lateralization, as indexed bymixed handedness, is associated with mild sub clinical language dysfunction, rather than non-verbal cognitive ability, and thismight be relevant to the expression of psychosis phenotype.© 2006 Elsevier B.V. All rights reserved.

Keywords: Cerebral lateralization; Mixed handedness; Schizophrenia; Schizotypy; SPQ

1. Introduction

As a proxy measure of decreased cerebral lateraliza-tion, mixed handedness or ambitexrality has been

⁎ Corresponding author. Department of Psychiatry, National andKapodistrian University of Athens Medical School, Eginition Hospital74 Vas. Sofias Ave. Athens 11528, Greece. Tel.: +30 6932 658067;fax: +30 210 7242020.

E-mail address: nistefan@med.uoa.gr (N.C. Stefanis).

0920-9964/$ - see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.schres.2006.04.021

associated in many studies with schizophrenia andelevated schizotypy scores in non-clinical samples, thusreinforcing the notion that subtle disturbances in theestablishment of normal cerebral asymmetry, might tosome extent be related to the etiology of the disorder(Flor-Henry, 1969; Crow et al., 1989) and/or symptomexpression within the psychosis extended phenotype(Barnett and Corballis, 2002; Chapman and Chapman,1987; Claridge et al., 1998; Dollfus et al., 2002; Gregoryet al., 2003; Kim et al., 1992; Manoach et al., 1988;

290 N.C. Stefanis et al. / Schizophrenia Research 87 (2006) 289–296

Manoach, 1994; Poreh, 1994; Shaw et al., 2001;Sperling et al., 1999). Several studies however, havefailed to confirm the schizotypy–handedness relation-ship (Dinn et al., 2002; Dragovic et al., 2005; Jaspers-Fayer and Peters, 2005; Overby, 1993). Numerousconfounding factors may account for these discrepan-cies across studies ranging from different definitions andmethods of assessment of mixed handedness, todifferences in population characteristics. The primaryaim of this study was to examine if mixed handednessassessed by three different definitions of hand prefer-ence membership, is associated with the dimensionalendorsement of self rated schizotypal personality traitsin a large sample of young male conscripts who arerepresentative for the Greek population of this agestratum. Since it has been debated if cerebral laterali-zation processes may correlate with cognitive ability(Annett and Manning, 1989, 1990; Leask and Crow,2005) we wished to examine further if hand preferencediscriminates conscripts on non-verbal cognitiveperformance.

2. Methods

2.1. Participants

Participants were volunteers drawn randomly fromeight consecutive draft board presentations of youngconscripts aged 18–24 at the recruitment center of theAir force military base, as part of the ongoing ASPISstudy (Athens Study of Psychosis Proneness andIncidence of Schizophrenia; Stefanis et al., 2004a).Military service is compulsory in Greece and all healthymales are recruited and assigned to the different armycorps by random assignment. All conscripts had beenevaluated as not suffering from a major medicalcondition. The study was approved by The Bioethicsand Medical Deontology Committee of the UniversityMental Health Research Institute. The initial sampleconsisted of 1353 conscripts who had reliable completedthe Schizotypal Personality Questionnaire (SPQ; Raine,1991), of which 1129 (83.4%) demonstrated handpreference for the 6 primary actions derived from theAnnett (1970) Hand Preference Questionnaire (HPQ)and 947 conscripts (70%) demonstrated in addition,hand preference for actions derived from the Porac andCoren (1981) questionnaire of lateral preferences (P/C).

2.2. Handedness assessment

The 6 primary actions in the HPQ consist of handpreference for writing, throwing ball, hammering nail,

brushing teeth, striking match, and holding a racquet.The six secondary actions defined in the HPQ were notassessed. Responses were coded as “right”, “left”, or“either” to each action. Based on his response pattern tothe 6 primary actions, each conscript was assigned toone of three classes “right”, “left”, or “mixed” accordingto the following two definitions of hand preference.

A) In adherence to Annett's (1970, 1985) recom-mendation, responses of “either” were coded as thewriting hand and mixed handed individuals weredefined as anyone demonstrating right or eitherpreference for writing but left hand preference for anyother primary action (original class 5) and anyindividual demonstrating left or either preference forwriting but right hand preference for any other primaryaction (original class 6). We combined classes 5 and 6into a mixed handedness group. We therefore adopted anarrow definition of mixed handedness since it has beensuggested that this would best capture the “point ofhemispheric indecision” that Crow (1997) claims is acore feature of schizophrenia. Individuals belonging tothe original Annett classes 1–4 were grouped anddefined as right handed and individuals belonging to theoriginal Annett classes 7 and 8 were grouped anddefined as left handed.

B) We further opted for an alternative a-theoreticalbut data oriented exploratory approach for handpreference membership utilizing Cluster Analysismethodology (Hierarchical clustering, Average Linkagebetween groups comparison; Anderberg, 1973). Petersand Murphy (1992) have suggested that it is possible toproduce distinct groups of handedness by using clusteranalytical methods. The two initial left and right handedgroups were defined as in the previous categorization;the remaining 323 subjects were assigned to threegroups according to cluster analysis methodology. Thecentral group was defined as the mixed handed, whilethe other two groups were defined as right and lefthanded groups and therefore were added to thecorresponding initial groups.

C) Handedness was in addition, assessed by directdemonstration to the four actions (throwing ball,drawing, erasing and removing top card while dealing)derived from the P/C questionnaire. Each actiondemonstrated with the left hand was coded as −1, withthe right hand as 1 while actions using either hand werecoded by zero. The scores for each individual in all taskswere summed and divided by 4 (the number of tasks).The conscripts whose final scores were −1 and −0.5were coded as left handers, the ones whose scores were0.5 and 1 were coded as right and the ones with middlescores (−0.3 and 0.3) formed the mixed handed group.

291N.C. Stefanis et al. / Schizophrenia Research 87 (2006) 289–296

We also used a dichotomous definition of handednessby dividing individuals in two mixed and non-mixed(right and left) groups, based on the previous three handmembership definitions.

2.3. Schizotypy assessment

The SPQ is a 74-item questionnaire that assesses allnine aspects of the Schizotypal Personality Disorder(SPD) according the Diagnostic and Statistical Manualof Mental Disorders Revised (DSM-III-R; AmericanPsychiatric Association, 1987). The analysis of thefactorial structure of the SPQ as defined by theresponses of the Greek sample has been assessed thoughconfirmatory factor analysis and indicated that the bestfit to the data was provided by a four factor modelnamely Positive, Negative, Disorganization and Para-noid factor (Stefanis et al., 2004a).

2.4. Non-verbal cognitive ability assessment

Conscripts underwent an extensive interview ofcomputerized neurocognitive abilities including assess-ment of vigilance with the Continuous PerformanceTask-Identical Pair version (CPT-IP; Cornblatt et al.,1988, 1989), verbal and spatial working memory with aversion of N-Back (Gevins et al., 1996) and non-verbalIQ was estimated with Raven Progressive Matrices(RPM; Raven, 1988). For each of the three computer-ized cognitive task, the central index of performanceaccuracy (d′) was used in subsequent analyses asdescribed previously (Stefanis et al., 2004b). Ninehundred and twenty-five conscripts (81.2%) had validmeasures on all four cognitive tasks.

Table 1Descriptive indices of SPQ factor and SPQ total scores for the three groups

Method of classification Group 1. Positive 2. Negativ

Mean (S.D.) Mean (S.D

Annett (N=1129) Right 0.199 (0.11) 0.180 (0.1Mixed 0.224 (0.12) 0.198 (0.1Left 0.218 (0.11) 0.208 (0.1Total 0.202 (0.11) 0.183 (0.1

Cluster analysis (N=1129) Right 0.200 (0.11) 0.180 (0.1Mixed 0.223 (0.12) 0.201 (0.1Left 0.211 (0.11) 0.207 (0.1Total 0.202 (0.11) 0.183 (0.1

Porac–Coren (N=947) Right 0.199 (0.11) 0.184 (0.1Mixed 0.226 (0.12) 0.198 (0.1Left 0.209 (0.11) 0.187 (0.1Total 0.201 (0.11) 0.184 (0.1

a, b, c, d, eThe scores of the three groups within columns were significantly d⁎,¶The scores of the two groups were significantly different according Dunn

2.5. Statistical analysis

The normality assumption was rejected in all cases(Lilliefors, 1967; Kolmogorov–Smirnov Test of Nor-mality: p>0.05). Hence, the corresponding non-para-metric Kruskal–Wallis test (KW χ2) was used, followedby Dunn (1964) post hoc tests for pairwise comparisons.For the comparisons of the mixed with the non-mixedindividuals the Wilcoxon/Mann–Whitney (MW U) testwas used. The agreement between the 3 alternativememberships and the test–retest reliability of the HPQwere assessed by Cohen's (1960) Kappa coefficient. Alldata analyses were conducted on the SPSS 10.0statistical package.

3. Results

3.1. Comparison of schizotypy scores across handpreference membership

Using the Annett hand preference membership,based on the 6 primary actions, 974 conscripts(86.3%) were assigned in the right-handed group, 92conscripts (8.1%) were assigned in mixed group and 63conscripts (5.6%) were assigned to the left handedgroup.

When the Cluster analysis approach to handpreference membership was used, 1003 cases (88.8%)were assigned to the right handed group, 50 cases(4.4%) were assigned to the mixed handed group and 76cases (6.7%) to the left handed group.

Using the P/C criteria for hand preference member-ship, 839 (74.3%) individuals were classified as righthanders, 46 (4.1%) as mixed handers and 62 (5.5%) as

derived from each classification

e 3. Disorganization 4. Paranoid Total SPQ

.) Mean (S.D.) Mean (S.D.) Mean (S.D.)

2) a0.414 (0.20)⁎ 0.441 (0.19) b36.752 (16.73)2) a0.457 (0.19)⁎ 0.474 (0.18) b39.823 (16.06)2) a0.461 (0.20) 0.484 (0.17) b41.312 (16.44)2) 0.420 (0.20) 0.447 (0.18) 37.286 (16.69)2) c0.415 (0.20) 0.443 (0.18) d36.871 (16.73)2) c0.473 (0.19) 0.467 (0.18) d40.201 (16.42)2) c0.456 (0.19) 0.487 (0.17) d40.696 (16.03)2) 0.420 (0.20) 0.447 (0.18) 37.286 (16.69)2) e0.416 (0.20)¶ 0.444 (0.18) 37.012 (16.55)2) e0.483 (0.18)¶ 0.484 (0.17) 40.691 (15.36)1) e0.438 (0.19) 0.477 (0.17) 38.981 (16.16)2) 0.421 (0.20) 0.448 (0.18) 37.319 (16.48)

ifferent according Kruskal–Wallis test (p<0.05).post hoc test (p<0.017).

Table 2Kruskal–Wallis test of differences in the SPQ factor and total scores between the three handedness groups derived from the each hand preferenceclassification

Method of classification Test statistics 1. Positive 2. Negative 3. Disorganization 4. Paranoid Total SPQ

Annett (N=1129) KW ⁎ χ2 5.78 5.85 7.04 4.94 6.69df 2.00 2.00 2.00 2.00 2.00p-value 0.055 0.054 0.030 0.085 0.035

Cluster analysis (N=1129) KW χ2 2.91 5.45 7.55 4.80 6.06df 2 2 2 2 2p-value 0.233 0.066 0.023 0.091 0.048

Porac–Coren (N=947) KW χ2 3.06 1.17 6.32 3.80 3.23df 2.00 2.00 2.00 2.00 2.00p-value 0.216 0.558 0.042 0.150 0.198

⁎ Kruskal–Wallis Chi-Square.

292 N.C. Stefanis et al. / Schizophrenia Research 87 (2006) 289–296

left handers. The definition of mixed handednessaccording to Annet hand preference membership wasthe least stringent of the three, since only 46% and 41%of these individuals, were also characterized as mixedhanded by cluster analysis and P/C criteria respectively.

No differences in age, and years of educationemerged between the three handedness groups in eitherof three hand preference memberships. The degree ofagreement on class status between the three definitionsof hand preference assessed with Kappa statistics andfound by pair wise comparisons greater than 0.71indicating good and high agreement between the threedefinitions of hand preference membership (Landis andKoch, 1977).

Significant differences between the three handednessgroups according to the Annett hand preferencemembership emerged for the Disorganization SPQfactor and for the total SPQ score (Tables 1 and 2). Inthe cases of the other three factors the differences wereevidential of a trend (p<0.10), with the Positive andNegative factors being marginally significant (Table 2).The pairwise comparisons were also assessed using theDunn test (1964). In all comparisons, the significancelevel was set up to α=0.10 concluding to significancelevel α′=0.017 for each paired comparison. The Dunntest yielded significant difference in the case of thescores of the Disorganization factor between right and

Table 3Descriptive indices of SPQ factor and total scores between non-mixed and m

Method of classification Group 1. Positive 2. Neg

Annett (N=1129) Non-Mixed a0.200 (0.11) 0.181Mixed a0.225 (0.12) 0.198

Cluster analysis (N=1129) Non-mixed 0.201 (0.11) 0.182Mixed 0.223 (0.12) 0.201

Porac–Coren (N=947) Non-Mixed 0.200 (0.11) 0.184Mixed 0.226 (0.12) 0.198

a, b, c, dThe scores of the two groups within columns were significantly diffe

mixed handers (p=0.012) with mixed handed indivi-duals scoring higher (Table 1).

The comparisons of the SPQ schizotypy scoresbetween the three handedness groups using the clusteranalysis methodology (Tables 1 and 2), led us to similarconclusions in the cases of the Disorganization factorand total SPQ scores. However, Dunn test did not yieldany significant pairwise differences. The differences inthe scores between the three handedness groups weremarginally significant for the other SPQ factors (Table2).

Replicating previous analyses by applying the P/Cdefinition of hand preference membership, significantdifferences were observed for the Disorganizationfactor. None of the other SPQ factors (Positive,Negative, or Paranoid) or the total SPQ score differedsignificantly amongst the three handedness groupsdefined by P/C criteria.

At the 9 SPQ subscale level, ‘odd speech’ was thesubscale that was most persistently associated withsignificant differences (p<0.05) between the threehandedness groups (data not shown).

The mixed handedness group had significantlyhigher scores than the non-mixed, on the Disorganiza-tion SPQ factor in all dichotomous defined handednessgroups (Tables 3 and 4). Significant differences in thescores of the SPQ Positive factor among mixed and non-

ixed groups derived from the each hand preference classification

ative 3. Disorganization 4. Paranoid Total SPQ

(0.12) b0.417 (0.20) 0.444 (0.18) 37.017 (16.74)(0.12) b.457 (0.18) 0.474 (0.17) 39.818 (16.06)(0.12) c0.418 (0.20) 0.446 (0.18) 37.144 (16.70)(0.12) c0.473 (0.19) 0.467 (0.18) 40.204 (16.42)(0.12) d0.418 (0.20) 0.447 (0.18) 37.194 (16.53)(0.12) d0.483 (0.18) 0.484 (0.18) 40.691 (15.36)

rent according Wilcoxon/Mann–Whitney test (p<0.05).

293N.C. Stefanis et al. / Schizophrenia Research 87 (2006) 289–296

mixed handed groups was noted only using the Annetthand preference membership.

No significant differences between mixed and non-mixed handedness groups were observed for the totalSPQ score, the Negative and Paranoid SPQ factors.

3.2. Test–retest of Annett membership

Two years after the initial assessment, a sub-sampleof 260 conscripts was summoned to participate in aretest analysis including hand preference on the 6primary actions. The Kappa coefficient on classmembership was 0.784 (p<0.001), indicating highagreement between the two assessments of individual'shand preferences.

3.3. Non-verbal cognitive ability and hand preferencemembership

No differences in non-verbal cognitive ability werefound across the three handedness groups (right, mixed,left) according to the Annett hand preference member-ship (CPT-IP: KW χ2 =0.29, df=2, p=0.869; verbal N-Back: KW χ2 =0.66, df=2, p=0.717; spatial N-Back:KW χ2 =0.248, df=2, p=0.883; Raven PM scores: KWχ2 =0.64, df=2, p=0.726). Likewise, no differencesemerged when hand preference membership wasdefined either by cluster analysis or by P/C criteria.Additionally, mixed versus non-mixed handednessgroups did not differ in terms of non-verbal cognitiveability.

4. Discussion

Our results indicate that self rated schizotypalpersonality traits are sensitive to hand preferencewhich is considered an indirect marker for the degreeof cerebral lateralization. Significant differences of thetotal SPQ scores were found between the three groups ofhand preference defined by theoretical and data oriented

Table 4Mann–Whitney test of differences in the SPQ factor and total scores betweeclassification

Method of classification Group 1. Positive 2. Neg

Annett (N=1129) MW-U ⁎ 47,664.5 49,775p-value 0.040 0

Cluster analysis (N=1129) MW-U 23,711.5 24,440p-value 0.148 0

Porac–Coren (N=947) MW-U 17,860.0 18,981p-value 0.114 0

⁎ Mann–Whitney U.

approaches. Disorganization schizotypy was the mostsensitive SPQ dimension to hand preference. Thisdimension was consistently associated with mixedhandedness regardless of hand preference classification.The association was even more significant as morestringent criteria for mixed handedness were applied(Table 4). The Disorganization SPQ factor encompassesthe subscales of ‘odd speech’ and ‘odd behavior’, andcontains items related to self perceived difficulties inoral communication, and oddities in social behavior.The significant association between ambitextrality andoddities in verbal communication within apparentlyhealthy individuals supports the validity of previousreports according to which lateralization is a determi-nant of verbal abilities in the general population (Crow,1998; Leask and Crow, 2005). In these studies, equalhand skilled children had relatively more verbal deficits,analogous perhaps to the deviations in verbal commu-nication reported by ambidextrous conscripts in thisstudy.

The association between mixed handedness andDisorganization schizotypy reported here is reminiscentof the corresponding association between mixedhandedness and formal thought disorder in schizophre-nia (Dollfus et al., 2002; Manoach, 1994). It may beviewed therefore, under a dimensional perspectivewhere deviant cerebral lateralization, indexed by apreference for mixed handedness, interferes withlanguage related processes. This relationship may actin such a way as to bias ambidextrous individualstowards language related dysfunction across a gradientof severity, ranging from relative mild deficits in verbalabilities and sub clinical oddities in verbal communica-tion amongst apparently healthy individuals, as sug-gested in this study, to the manifestation of formalthought disorder in schizophrenia. According to thishypothesis, mixed handedness may constitute a riskfactor for language related disorder across the spectrum(from schizotypy to schizophrenia) and is not necessar-ily related specifically to schizophrenia per se. Indeed it

n non-mixed and mixed groups derived from the each hand preference

ative 3. Disorganization 4. Paranoid Total SPQ

.5 47,367.5 48,905.5 48,830.5

.164 0.032 0.096 0.083

.5 21,910.5 24,568.5 23,383.5

.261 0.025 0.286 0.105

.0 16,369.0 17,981.0 17,816.5

.336 0.016 0.130 0.103

294 N.C. Stefanis et al. / Schizophrenia Research 87 (2006) 289–296

has been observed that lateralization problems incur riskfor language delays and deficits including stuttering(Orton, 1937); as well as reading, spelling and drawingdifficulties (Zangwill, 1960). Furthermore, mixed hand-edness has been associated with an increased risk forother language related disorders such as dyslexia(Richardson, 1994).

A bolder approach that views schizophrenia as adevelopmentally induced failure of lateralization of thecomponents of language would adopt a slightly differentinterpretation of our finding. Mixed handedness anddeficits in verbal abilities are viewed as combinedmarkers of an underlying causal process, that is, failureof the hemispheric dominance for language, since theyare over-represented in the disorder and in childrendestined to become ill (Crow, 1998; Leask and Crow,2005). This hypothesis would imply that mixed handedindividuals, high on Disorganization schizotypy wouldbe at an increased risk to develop psychosis. Due to the2-year prospective design of ASPIS, this hypothesiscannot be tested.

Ambidextrality was found to be associated toDisorganization rather than the Positive aspects ofschizotypy, in contrast to several previous studies (Kimet al., 1992; Poreh et al., 1997; Shaw et al., 2001).Following the Annett hand preference classification,mixed handed individuals scored higher (than non-mixed) on Disorganization as well as positive aspects ofschizotypy. This was the only significant associationwith positive schizotypy and hand preference in ouranalysis. Furthermore, and as opposed to Disorganiza-tion schizotypy, observed trends of association betweenpositive and negative SPQ factors with mixed handed-ness disappeared, when more stringent definitions ofmixed handedness were employed (by cluster analysisor P/C criteria) (Tables 2 and 4). It is therefore plausiblethat previous reported inconsistencies are to an extentdependent on broader definitions of mixed handedness.Moreover, a close inspection of previous work pub-lished on the association between mixed handednessand SPQ by Kim et al. (1992), Poreh et al. (1997)indicate that the strongest associations noted were withsubscales ‘odd speech’, and ‘odd behavior and appear-ance’ respectively, both components of the Disorgani-zation, rather than the positive schizotypy factor.

Although samples of students used in previousstudies may reflect selected individuals due to higheracademic achievement, cognitive ability and socialbackground, caution should be exercised since the onestudy to date to assess schizotypy (measured with theSPQ) in a representative sample of the generalcommunity failed to show an association with mixed

handedness (Dragovic et al., 2005). The focus on a malepopulation may have contributed to the strengthening ofthe association between schizotypy and atypical hand-edness as described in previous studies (Gregory et al.,2003; Shaw et al., 2001).

Finally, mixed handedness was not associated withdecreased performance on non-verbal cognitive tasks,supporting thus the notion that reduced cerebrallateralization, as indexed by mixed handedness, ismore sensitive to subtle verbal manifestation of languageprocesses rather than non-verbal cognitive ability.

5. Limitations

Handedness was assessed with demonstration onhand preference for the six primary actions and not forthe six secondary actions defined by the 12 item AHQscale, making comparability difficult with other studieson hand preference. We argue that in the full 12 itemAHQ scale, the core mixed handedness groups (classes5 and 6) that best capture the “point of hemisphericindecision” (Crow, 1997) are defined by the 6 primaryactions alone, with no need to incorporate the secondaryones. Culturally enforced writing with the right hand inotherwise left handed individuals was not controlled for.Nevertheless, the lengthier the tasks performed, thehigher the probability of establishing a task performedwith the opposite hand from writing, a fact that mayaccount for a high distribution of mixed handedness(30% of the population) when the 12 item questionnaireis used (Annett, 1972).

The predictive validity of schizotypy self rated scaleshas been questioned (Kendler et al., 1996). Moreover, ithas been proposed that the Disorganization schizotypyfactor encompassing oddities in speech and behaviorcannot be self rated reliably and is frequently under-reported (Bergman et al., 1996; Kendler et al., 1996;Vollema and Ormel, 1999). However, such a generalunder-reporting would have led to a bias towards thenull rather than induce mixed non-mixed handednessdifferences. Thus, the assumption of under-reportingwould confirm our results, unless one assumes thatmixed handed conscripts were more self aware andaccurate in reporting their behavioral and verbalcommunication difficulties compared to the non-mixedhandedness group. Although it is unlikely that apresumable heightened self awareness would biasmixed handers to report more verbal related difficultiesthan say paranoid beliefs or interpersonal difficulties,the army recruitment environment, which demandsquick responses to right–left commands, may confusemixed handers more than non-mixed handers.

295N.C. Stefanis et al. / Schizophrenia Research 87 (2006) 289–296

Acknowledgments

This work was supported by the grant “EKBAN 97"to Professor C.N. Stefanis from the General Secretariatof Research and Technology of the Greek Ministry ofDevelopment. The technical support for this project wasprovided by “Intrasoft Co”. We would like to thank thefollowing colleagues, in alphabetical order, who helpedin data acquisition and preprocessing: Ioannis Giouzelis,Georgios Kastrinakis, Emanouil Kattoulas, CatherinePaximadis, Christos Theleritis. We would also like tothank M. Annett for her valuable suggestions.

References

American Psychiatric Association, 1987. Diagnostic and StatisticalManual of Mental Disorders, 3rd edn. APA, Washington, DC.DSMR-III.

Anderberg, M.R., 1973. Cluster Analysis for Applications. AcademicPress, New York.

Annett, M., 1970. A classification of hand preference by associationanalysis. Br. J. Psychol. 61, 303–321.

Annett, M., 1972. The distribution of manual asymmetry. Br. J.Psychol. 63, 343–358.

Annett, M., 1985. Left, Right, Hand and Brain: The Right ShiftTheory. Lawrence Erlbaum, London.

Annett, M., Manning, M., 1989. The disadvantages of dextrality forintelligence. Br. J. Psychol. 80, 213–226.

Annett, M., Manning, M., 1990. Reading and a balancedpolymorphism for laterality and ability. J. Child Psychol.Psychiatry 31, 511–529 (Erratum in: 1994, J. Child Psychol.Psychiatry 35 573–575).

Barnett, K.J., Corballis, M.C., 2002. Ambidexterity and magicalideation. Laterality 7, 75–84.

Bergman, A.J., Harvey, P.D., Mitropoulou, V., Aronson, A., Marder,D., Silverman, J., Trestman, R., Siever, L.J., 1996. The factorstructure of schizotypal symptoms in a clinical population.Schizophr. Bull. 22 (3), 501–509.

Chapman, J.P., Chapman, L.J., 1987. Handedness of hypotheticallypsychosis-prone subjects. J. Abnorm. Psychology 96, 89–93.

Claridge, G., Clark, K., Davis, C., Mason, O., 1998. Schizophreniarisk and handedness: a mixed picture. Laterality 3, 209–220.

Cohen, J., 1960. A coefficient of agreement for nominal data. Educ.Psychol. Meas. 20, 37–46.

Cornblatt, B.A., Risch, N.J., Faris, G., Friedman, D., Erlenmeyer-Kimling, L., 1988. The Continuous Performance Test, identicalpairs version (CPT-IP): new findings about sustained attention innormal families. Psychiatry Res. 26, 223–238.

Cornblatt, B.A., Lenzenweger, M.F., Erlenmeyer-Kimling, L., 1989.The continuous performance test, identical pairs version: II.Contrasting attentional profiles in schizophrenic and depressedpatients. Psychiatry Res. 29, 65–85.

Crow, T.J., 1997. Is schizophrenia the price that Homo sapiens pays forlanguage? Schizophr. Res. 28, 127–141.

Crow, T.J., 1998. Sexual selection, timing and the descent of man: atheory of the genetic origins of language. Curr. Psychol. Cogn. 17,1079–1114.

Crow, T.J., Ball, J., Bloom, S.R., Brown, R., Bruton, C.J., Colter, N.,Frith, C.D., Johnstone, E.C., Owens, D.G., Roberts, G.W., 1989.

Schizophrenia as an anomaly of development of cerebralasymmetry. A postmortem study and a proposal concerning thegenetic basis of the disease. Arch. Gen. Psychiatry 46, 1145–1150.

Dinn, W.M., Harris, C.L., Aycicegi, A., Greene, P., Andover, M.S.,2002. Positive and negative schizotypy in a student sample:neurocognitive and clinical correlates. Schizophr. Res. 1 (56),171–185.

Dollfus, S., Buijsrogge, J.A., Benali, K., Delamillieure, P., Brazo, P.,2002. Sinistrality in subtypes of schizophrenia. Eur. Psychiatr. 17(5), 272–277.

Dragovic, M., Hammond, G., Jablensky, A., 2005. Schizotypy andmixed-handedness revisited. Psychiatry Res. 136, 143–152.

Dunn, J.O., 1964. Multiple comparisons Using Rank Sums. Techno-metrics 6, 241–252.

Flor-Henry, P., 1969. Psychosis and temporal lobe epilepsy: acontrolled investigation. Epilepsia 10, 365–395.

Gevins, A.S., Smith, M.E., Le, J., Leong, H., Bennett, J., Martin,N., McEvoy, L., Du, R., Whitfield, S., 1996. High resolutionevoked potential imaging of the cortical dynamics of humanworking memory. Electroencephalogr. Clin. Neurophysiol. 98,327–348.

Gregory, A.M., Claridge, G., Clark, K., Taylor, P.D., 2003.Handedness and schizotypy in a Japanese sample: an associationmasked by cultural effects on hand usage. Schizophr. Res. 15 (65),139–145.

Jaspers-Fayer, F., Peters, M., 2005. Hand preference, magical thinkingand left–right confusion. Laterality 10, 183–191.

Kendler, K.S., Thacker, L., Walsh, D., 1996. Self-report measures ofschizotypy as indices of familial vulnerability to schizophrenia.Schizophr. Bull. 22, 511–520.

Kim, D., Raine, A., Triphon, N., Green, M.F., 1992. Mixed-handedness and features of schizotypal personality in a non-clinical sample. J. Nerv. Ment. Dis. 180, 133–135.

Landis, J.R., Koch, G.G., 1977. The measurement of observeragreement for categorical data. Biometrics 33, 159–174.

Leask, S.J., Crow, T.J., 2005. Lateralization of verbal ability in pre-psychotic children. Psychiatry Res. 15 (136), 35–42.

Lilliefors, H.W., 1967. On the Kolmogorov–Smirnov test fornormality with mean and variance unknown. Jasa 64, 399–402.

Manoach, D.S., 1994. Handedness is related to formal thoughtdisorder and language dysfunction in schizophrenia. J. Clin. Exp.Neuropsychol. 16 (1), 2–14.

Manoach, D.S., Maher, B.A., Manschreck, T.C., 1988. Left-handed-ness and thought disorder in the schizophrenias. J. Abnorm.Psychology 97, 97–99.

Orton, S.T., 1937. Reading, Writing and Speech Problems in Children.Chapman & Hall, London.

Overby, L.A., 1993. Handedness patterns of psychosis-prone collegestudents. Pers. Individ. Differ. 15, 261–265.

Peters, M., Murphy, K., 1992. Cluster analysis reveals at least three,and possibly five, distinct handedness groups. Neuropsychologia30, 373–380.

Porac, C., Coren, S., 1981. Lateral Preferences and Human Behaviour.Springer-Verlag, New York.

Poreh, A., 1994. Reexamination of mixed handedness in psychosis-prone college students. Pers. Individ. Differ. 17, 445–448.

Poreh, A.M., Levin, J., Teves, H., States, J., 1997. Mixed handednessand schizotypal personality in a non-clinical sample. The role oftask demand. Pers. Individ. Differ. 23, 501–507.

Raine, A., 1991. The SPQ: a subscale for the assessment of schizotypalpersonality based on DSM-III-R criteria. Schizophr. Bull. 17 (4),555–564.

296 N.C. Stefanis et al. / Schizophrenia Research 87 (2006) 289–296

Raven, J.C., 1988. Raven's Progressive Matrices. Psychological Corp(Harcourt), San Antonio, TX.

Richardson, A.J., 1994. Dyslexia, handedness and syndromes ofpsychosis-proneness. Int. J. Psychophysiol. 18, 251–263.

Shaw, J., Claridge, G., Clark, K., 2001. Schizotypy and the shift fromdextrality: a study of handedness in a large non-clinical sample.Schizophr. Res. 50, 181–189.

Sperling, W., Martus, P., Barockam, A., 1999. Non-right-handednessand obstetrical complications in paranoid hallucinatory schizo-phrenics. Psychopathology 32, 267–276.

Stefanis, N.C., Smyrnis, N., Avrampoulos, D., Evdokimidis, I.,Ntzoufras, I., Stefanis, C.N., 2004a. Factorial composition of selfrated schizotypal traits amongst young males undergoing militarytraining. Schizophr. Bull. 30 (2), 335–350.

Stefanis, N.C., Van Os, J., Avramopoulos, D., Smyrnis, N.,Evdokimidis, I., Hantoumi, I., Stefanis, C.N., 2004b. Variation incatechol-o-methyltransferase val158 met genotype associated withschizotypy but not cognition: a population study in 543 youngmen. Biol. Psychiatry 56, 510–515.

Vollema, M.G., Ormel, G., 1999. A multitrait–multi method analysisof three dimensions of schizotypy. In: Vollema, M.G. (Ed.),Schizotypy: Toward the Psychological Heart of Schizophrenia.Shaker Publishing, Maastricht, Netherlands, pp. 87–102.

Zangwill, O.L., 1960. Cerebral Dominance and its Relation toPsychological Function. Oliver & Boyd, Edinburgh.