Do Leadership behavior, Team stability and Task interdependence relate to Team learning? 1
Transcript of Do Leadership behavior, Team stability and Task interdependence relate to Team learning? 1
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
1
Do Leadership behavior, Team stability and Task interdependence relate to
Team learning? 1
Chantal Savelsbergh, MSc (Corresponding author)
Open University of the Netherlands
P.O. Box 2960
6401 DL Heerlen, the Netherlands
Tel: +31 (0)45 576 28 26 / +31 (0) 6 46 23 60 60
Fax: +31 (0)45 576 21 03
E-mail: [email protected]
Prof. Peter M. Storm, PhD
Open University of the Netherlands
Kennis&Co b.v.
E-mail: [email protected]
Dr. Ben Kuipers, PhD
Erasmus University
E-mail: [email protected]
Submission type: Empirical Research Paper
Keywords: Team learning Behaviors, Leadership behavior, Team stability, Task
interdependence
1 This research has been funded partly by Project Management Institute America (PMI America)
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
2
Do Leadership behavior, Team stability and Task interdependence relate to Team learning?
Chantal Savelsbergh, Open universiteit Nederland, Heerlen Peter Storm, Open universiteit Nederland, Heerlen
Ben Kuipers, Erasmus Universiteit, Rotterdam
In professional as well as academic publications (Banker, Field, Schroeder, & Sinha, 1996;
Drucker, 2003; Glassop, 2002; Pfeffer, 1994), the importance of teamwork to organizational
success in today’s economy is emphasized continually. Teams are expected to enable
increased adaptability, productivity, and creativity compared to what individual employees
can offer. They are also believed to provide more innovative and comprehensive solutions to
complex organizational problems (cf. Beers, 2005). However, both practice and empirical
research demonstrate that there is considerable variance in team performance (e.g., Hackman,
1987).
The teams we address in this study are project teams in knowledge-intensive
organizations (cf. Starbuck, 1992). Many knowledge-intensive work settings are characterized
by overload, ambiguity and politics, where project teams face a multitude of possible
problems to address and solutions with which to address them. There is little way to know
which problems and solutions to select, and there exist multiple stakeholders with an interest
in their selections (Alvesson, 2004). The most important performance outcome for these
teams is the quality of the product they deliver to their clients. The work consists primarily of
gathering information, know-how, and feedback through interpersonal exchanges or reading,
analysing and applying this knowledge, and delivering written and oral conclusions to
colleagues and clients. This raises the expectation that team learning behaviors should be
especially beneficial for project teams in this kind of organizations, which is consistent with
team learning theories that argue that collective exploration, construction of meaning,
reflection and experimentation enhance team performance (e.g., Edmondson, 1999; Argote,
Gruenfeld, & Naquin, 2001). Previous research has singled out team learning behaviors as
one of the main influencing factors of team performance (e.g., Edmondson, 1999; Argote,
Gruenfeld, & Naquin, 2001; Savelsbergh, Van der Heijden, & Poell, 2007). Given the fluid
environments that project teams work within continuous learning is a key driver of a team’s
ability to remain adaptive and flexible. Remains the question of how of team learning
behaviors in teams can be promoted. And, how this can be done by the team itself. The aim of
this study was to single out antecedents of team learning behaviors in project teams, that can
be influenced by the team and their leader. The question is how the team leader and the team
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
3
can create conditions (e.g. psychological safety, Edmondson, 2004) that promote team
learning behaviors.
In this study we focus on four team learning enhancing antecedents: (1) leadership behavior,
(2) task interdependence and (3) fulltime vs part-time involvement in the team and (4) team
stability.
Theory
A team can be defined as “a distinguishable set of two or more people who are
assigned specific roles or functions to perform dynamically, interdependently, and adaptively
toward a common and valued goal/object/mission, who have each been assigned specific roles
or functions to perform, and who have a limited life span of membership” (Salas, Dickinson,
Converse, & Tannenbaum, 1992, p. 126). Additionally, project teams are characterized by a
unique goal and a planned start and ending.
Although much research has been done among samples of teams and on the topic of
learning in organizations, relatively little is known about team learning (c.f. Edmondson,
1999). In defining the concept of team learning, some researchers have emphasized the
process of learning (e.g., Edmondson, 1999, 2002; Gibson & Vermeulen, 2003; Kasl, Marsick
& Dechant, 1997), while others have stressed its outcomes (e.g., Ellis, Hollenbeck, Ilgen,
Porter, West & Moon, 2003). We follow the first stream and adhere to the rich definition of
team learning adopted by Edmondson (1999). Edmondson (1999) refers to various team
learning behaviors. In this study team learning is defined as an ongoing process of collective
reflection and action characterized by: (1) exploring; (2) co-construction of meaning, (3)
reflecting on outcomes & (4) processes; (5) communicating & (6) discussing errors and
unexpected outcomes of actions; (7) seeking feedback; and (8) experimenting within and as a
team (Savelsbergh, van der Heijden & Poell, 2008). Descriptions of these team learning
behaviors are summarized in Table 1.
**** Insert Table 1 about here ****
Teams differ in the extent to which they engage in learning behaviors (e.g.,
Edmondson, 1999). It has been shown that a positive relationship exists between team
learning behaviors and team performance (Gibson & Vermeulen, 2003; Van der Vegt &
Bunderson, 2005). How can this variance in learning behaviors be explained? In this study we
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
4
focus on those antecedents of team learning behavior that may be influenced by the team
members and leader themselves.
First, we were interested in how the leader’s behavior relates to team learning.
Fleishman and colleagues (1991) noted person-focused and task-focused as the most common
classification system in literature on individual and team leadership behavior. Person-oriented
leadership behaviors facilitate team interaction and/or development. Task-oriented leadership
behavior consists of behaviors which work to ensure that members have a clear sense of
direction and purpose, that guide the team action toward goal attainment. Joining this
traditional dichotomy we focus on four person-focused and one task-focused leadership
behaviors, that are expected to be effective in work teams (Stoker, 1998, p.58). Our aim is to
single out which of these leadership behaviors relate to learning behavior in teams and as such
have an additional indirect relationship to on team performance.
The person-focused leadership behaviors are Consideration, Coaching, Consultative or
Participative and Charismatic leadership behaviors. The task-focused leadership behavior is
Initiating Structure. Consideration as a person-focused and Initiating Structure as a task-
focused leadership behavior (Burke, 2006; Judge, Piccolo, et al., 2004; Stogdill, 1950) have
proven to be among the most robust leadership concepts. Stoker (1998) added Coaching,
Consultative and Charismatic leadership as contemporary person-focused leadership
behaviors that were expected to have influence on the effectiveness of self-managed work
teams. These four person-focused behaviors all differ in their focus and as such are expected
to have a complementary relationship with effectiveness. Considerate leaders show concern
and respect for followers, look for their welfare, and give support. Coaching is defined as day-
to-day encouragement of improving employees’ own performance (Popper and Lipshitz,
1992). Consultative or Participative leadership refers to involvement in decision making
(House & Mitchell, 1974). And Charismatic leaders demonstrate behavior, that shows their
powerful personality and vision and are trusted and respected by their subordinates (Bass,
1990a & 1990 b).
Findings from previous studies confirm that person-focused leadership behavior relate
to team learning. Burke and collegues (2006) demonstrated that this behavior explains 31% of
variance in team learning outcome. Research by Edmondson (1996, 1999) shows that
coaching team leaders giving guidance, encouragement and support to the team members
influence psychological safety in teams, which in turn promotes team learning behaviors.
Based on these findings, we speculate that all four person-oriented leadership behaviors
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
5
promote team learning behavior. In addition, we expect that the relationships of these four
person-focused leadership behaviors complement each other in explaining variance of team
learning. In other words, we expect that a mixture of these four leadership behaviors, has a
stronger influence on the extent of team learning compared to the influence of each separate
person-focused leadership behavior. These expectations lead to the following hypotheses:
Hypothesis 1a:
Consideration, Coaching, Consultative/Participative, as well as Charismatic leadership
behavior (Person-oriented leadership behaviors) are positively related to team learning
behaviors.
Hypothesis 1b:
A mixture of Consideration, Coaching, Consultative/Participative, and Charismatic
leadership behavior explains more variance in team learning behavior, as compared to
the explained variance by each separate person-focused leadership behavior.
As it comes to task-focused leadership, there is lack of evidence from previous research about
its relationship to team learning. Finding, however, demonstrate a strong influence on team
performance (.23) (Judge, Piccolo, et al., 2004). This pleads for not neglecting these task-
focused leadership behaviors and get an answer on the question if task-focused leadership
relates to team learning processes and as such has an indirect relationship to team
performance. As a rivaling hypothesis we expect that task-focused leadership negatively
related to team learning because it may frustrate the self-management potential of a team
(Stewart & Manz, 1995). These arguments lead to the following hypothesis:
Hypothesis 2:
Task-oriented leadership behavior is negatively related to team learning behaviors.
From a practical point of view, outcomes will help project managers to develop a mix of
leadership behaviors that is most helpful in developing team learning behaviors in their teams.
Second, we were interested whether team stability relates to team learning.
Membership change is low in a stable team, whereas turnover is high in an unstable team.
Research on team stability has emphasized the disruption caused by member turnover on
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
6
functioning and project performance due to knowledge depreciation. For instance, when
studying 211 NPD projects, Akgün and Lynn (2002) found that team stability positively
relates to team learning and project success. But the relationship of team stability with team
learning and performance is a matter of some debate in the literature (Edmondson, Bohmer &
Pisano, 2001). On the one hand, keeping the same team together facilitates coordination of
interdependent work. Experimental research has shown that keeping team members together
helps them to understand one another’s capabilities and to coordinate their actions (Moreland
and Levine, 1989). So they might become more capable of coordinating collective learning
behavior. On the other hand, over time, stable teams may become slaves to routine and fail to
respond to changing conditions.
Edmondson’s et al. (2007) report in their review on team learning that teams with a more
stable composition demonstrate higher rates of improvement. Especially when it comes to
learning by doing, as is the case with team learning behavior, they claim team stability as an
influencing factor.
Given the fact that project teams have a temporally lifespan with a unique assignment, the
risks of excessive stability in which the same team members are together for years facing the
same job are small. Therefore we anticipate that team stability could facilitate collective
learning in project teams.
Hypothesis 2:
Team stability is positively related to team learning behaviors.
Third, the relationship between task interdependence and team learning behaviors will
be examined. Task interdependence refers to the degree to which team members must depend
upon one another to perform their tasks in route to goal accomplishment (Saavedra, Earley, &
Van Dyne, 1993). As the level of task interdependence increases, so do the dependencies
amongst team members and thereby the coordination requirements needed to achieve
efficacious performance outcomes. Studies have shown that task interdependence leads to
more communication, helping, and information sharing than individualistic tasks (Crawford &
Haaland, 1972; Johnson, 1973). Wageman (1995) concludes that high task interdependence is
critical, whenever collaborative behavior is important, because they seem to lead to a shared
responsibility on the team level. So one may conclude that high task interdependence
promotes team learning behaviors, being all examples of collaborative behavior. This
expectation is supported by findings of Van den Bossche et al. (2006), that show that beliefs
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
7
about task and outcome interdependence seem to promote team learning processes. In this
study we focus on the relationship of perceived task interdependence on team learning.
Hypothesis 3:
Task interdependence is positively related to team learning behaviors.
We also will explore the relative strength of the relationship between these antecedents
(leadership behaviors, team stability and task interdependence) and team learning. Outcomes
may help project managers to focus interventions to promote team learning.
Method
Subjects and procedure
The investigation which is reported in this paper consists of a cross-sectional study among 41
project teams in the areas of Engineering, Building, Infrastructure and Information
Technology. The main activities undertaken by the project teams concern the design,
development and bringing to life of large ICT-systems, utilities or infrastructure. The survey
was sent to all the members of the participating teams and their project manager (391 in
total). Responses were received from 274 team members and project managers, giving an
response rate of 70%. The mean age was 42.65 years (DS= 10.11), and 84,7% of the
repondents were male.
Measures
A questionnaire was designed to measure team learning behaviors, leadership behaviors, team
stability and task interdependence. The survey was pretested in face-to-face interviews with 3
individuals and completed in part or full by 4 others before it was finalized.
All scales covering leadership behaviors and task interdependence have been derived from
previously developed and validated instruments. New scales were developed for team
learning behaviour. These have been tested in a validation study among nineteen teams
(Savelsbergh et.al., 2008). Unless otherwise noted, five-point Likert-type scales ranging from
1 (completely disagree) to 5 (completely agree) were used in this study.
Initiating structure leadership behavior was measured using 2 items from the scale developed
by Stoker (1998) (based on the Ohio-State leadership questionnaire of Stogdill, 1963; Mulder,
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
8
Ritsma van Eck, de Jong, 1971). An example item is ‘My project manager quickly takes
charge’. The alpha coefficient of this scale was .65.
Consideration leadership behavior was measured using 2 items from the scale developed by
Stoker (1998) (based on the Ohio-State leadership questionnaire of Stogdill, 1950; Mulder,
Ritsma van Eck, de Jong, 1971). An example item is ‘I feel appreciated by my project
manager’. The alpha coefficient of this scale was .70.
Coaching leadership behavior was measured using 2 items from the scale developed by
Stoker (1998) (based on De Jong and Carpay (1991). An example item is ‘my project
manager gives me advice when I need it’). The alpha coefficient of this scale was .79.
Consultative or Participative leadership behavior using 2 items from the scale developed by
Stoker (1998) (based on Mulder, De Jong, Koppelaar & Verhage, 1986; Le Blanc, 1994). An
example item is ‘My project manager delegates enough responsibilities to team members’.
The alpha coefficient of this scale was .60.
Charismatic leadership behavior was measured using 2 items from the scale developed by
Stoker (1998) (based on Bass, 1985; Den Hartog et al., 1994). An example item is ‘My
project manager serves as an example to me’. The alpha coefficient of this scale was .55.
Although this was less than .60, we decided not to skip this variable, but to be cautious in our
conclusions.
Team stability was measured using two self constructed items to count the events how often
team member composition changed in the last year. The items are ‘How many persons newly
joined the project team in the last 12 months’ and ‘How many persons left the project team in
the last 12 months’. Responses on these two items are accumulated and used to categorize
teams into high stable teams (number of membership changes = 0), medium stable teams
(number of membership changes = 1-4) and low stable teams (number of membership
changes = 5 or more). The mean number of team members in the project teams was 9.5.
Task interdependence was measured using 3 items from Campion, Medsker & Higgs (1993).
An example item is ‘I cannot accomplish my tasks without input from other members of my
team’. The alpha coefficient of this scale was .78.
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
9
Team learning behavior was measured using 28 items developed by Savelsbergh (2008)
(based on van den Bossche, 2006; Van Dyck, 2000; Van Dyck, Frese, Baer, & Sonnentag
(2005), Edmondson, 1999; Schippers et al., 2003; Van Woerkom, 2003). This measurement
instrument is broken down into 8 dimensions (see for more information Table 1). An example
item is ‘Team members elaborate on each other’s information and ideas’. The alpha
coefficient of this scale was .94.
Analysis
All the variables were measured and analysed at the individual level. This means that the
outcomes reflect to what extent individually perceived leadership behavior, team stability and
task interdependence relate to the individually perceived extent of team learning behaviors.
Data analysis comprised several stages.
First, data screening was conducted to identify and to establish: (a) missing data; (b)
univariate normality and potential outliers; and (c) bivariate linearity, normality, and potential
outliers associated with the hypothesized correlations. Linear regression plots were examined
in order to test whether the assumptions were violated, which appeared not to be the case.
Second, a Confirmatory Factor Analysis (CFA) (Arbuckle, 2006) was performed to
test whether the measurement instrument for team learning behaviors could be used as a
suitable model. CFA and further hypothesis testing analyses have been performed using the
AMOS 16.0, a Structural Equation Modeling (SEM) program (Arbuckle, 2006). SEM was
chosen specifically for this investigation because of its capacity to handle complex models
with measurement error, and observed and latent variables. By explicitly estimating and
isolating the measurement error in observed variables, SEM reveals true variance and its
related effects upon variables in a model (Aragon & Gesell, 2003). Moreover, it facilitates
testing whether the hypothesized model fits, that is to say, whether it is supported by the
empirical data.
Third, we examined the relationship between the independent variable (team learning
behaviors as perceived by the team members and their leaders), and each separate dependent
variables (the five separate leadership behaviors, team stability and task interdependence as
perceived by the team members and project managers) using SEM (Arbuckle, 2006).
Fourth, we tested the relationship between the independent variable (team learning
behaviors) and the four person-focused leadership behaviors (Consideration, Coaching,
Consultative and Charismatic leadership behaviors) together in one model using SEM.
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
10
And finally, we tested in a combined model the relationships between the independent
variable (team learning behaviors) and the antecedents that appeared to be significantly
related to team learning behaviors in our third step.
Results
Descriptive Measures
Data collection took place from June to September 2008.
Table 2 presents the means, standard deviations, and correlations among all variables under
study. It also contains information on scale reliabilities and number of items.
**** Insert Table 2 about here ****
In order to test the fit between a SEM model and the data, the traditional Chi-squared value,
the goodness-of-fit index (GFI), and the sample root mean square error of approximation
(RMSEA) were calculated. As a rule of thumb, a GFI ≥ .90 and a RMSEA ≤ .08 indicate a
reasonable fit between the model and the data (Browne & Cudeck, 1993). Additionally,
PCLOSE, which is a p-value for testing the null hypothesis that the population RMSEA is no
greater than .05, has been determined (Arbuckle, 2006). In case of a RMSEA ≥ .05, the null
hypothesis is rejected indicating a lack of close fit. Because of the fact that the GFI and the
RMSEA are dependent upon size, as recommended by Marsh, Balla, and Hau (1996), the
non-normed fit index (NFI), and the comparative fit index (CFI) were also examined. These
indices should have values of .90 or higher (Hoyle, 1995).
The second-order confirmatory factor analysis (CFA) for the eight dimensions of team
learning behaviors (using the team learning behaviors instrument of Savelsbergh (2008)
consisting of 28 items) indicated a single second-order factor solution with an acceptable fit
(Chi square = 470.787, df = 295, p = 0.000; NFI = .885, CFI = .953, RMSEA = .047,
PCLOSE = 0.749). See Appendix 1 for more specific information.
Table 3 presents the outcomes of the examination of the relationship between each of
the five separate leadership behaviors, team stability and task interdependence and the team
learning behaviors.
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
11
**** Insert Table 3 about here ****
From the leadership behaviors each of the person-focused leadership behaviors appeared to be
significantly related to team learning. Our fourth step resulted in a model that visualizes the
relationships between these four person-focused leadership behaviors together and learning
behaviors (see Figure 2).
**** Insert Figure 1 about here ****
For sake of readability, the covariances, individual items with errors are not depicted.
One can see that 17 % of variance in team learning behaviors is explained by person-focused
leadership behaviors. However, in this mixture of person-focused leadership behaviors only
two behaviors appear to be significantly positively related to team learning, namely,
Charismatic leadership behavior (i.e., β = .28 p = .000) and Consultative or Participative
leadership behavior (i.e., β = .14, p = .047).
Finally, Figure 3 visualizes the outcomes of the examination of the relationships from
each of the four significant leadership behaviors, team stability and task interdependence to
the team learning in one model.
**** Insert Figure 2 about here ****
As is clear from Figure 3, 18 % of variance in team learning is explained by these
combination of expected antecedents. Only two antecedents remained significant in this
combined model, viz. Charismatic leadership (i.e., β = .28 p = .000) and Team Stability (i.e.,
β = .14 p = .022).
Conclusions and Discussion
Based on the outcomes of our study we draw several conclusions. (1) each person-
focused leadership behavior relates significantly and positively to learning behavior in project
teams (see Table 3). (2) Task-focused leadership does not relate to learning behavior in teams.
(3) All person-focused leadership behaviors taken together explain more variance in learning
(17%) compared to their separate explaining power (see Table 3). Moreover, in this
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
12
combination of leadership behaviors only Consultative and Charismatic leadership behavior
remain significantly related to learning. (4) Team stability as well as (5) Task interdependence
relates significantly and positively to team learning. (5) A combined model of all antecedents
of team learning in this study explains more variance in learning (18%) compared to the
explanatory power of the separate antecedents (see Table 3) or the mix of person-focused
leadership behaviors (17%).
The relationships between the various person-focused leadership behaviors and
learning seem to differ in their relative strength. As the findings demonstrate: a project
manager using a mix of leadership behaviors promotes team learning particularly through
charismatic and consultative leadership behavior.
How can we explain these findings? Consultative leadership involves team members
in decision making. It challenges the responsibility of all team members. It might be that
project managers stimulate team learning by example. By listening to the perspectives of team
members and combining their views to a collective view.
Charismatic leadership refers to having a vision and living by it. It might be the case
that this behavior helps team members to get a sense of direction, without filling in every
detail about how to get there. Not everything is worked out into tasks and procedures, as is
more the outcome of task-focused leadership behavior. The team members still need to co-
construct their collective perspective on this vision and how they want to work it out together.
This might evoke team learning with a clear sense of purpose and direction. A charismatic
project manager might make an example by showing team learning behaviors himself. He
shows self confidence in what he is standing for by communicating a clear and appealing
vision. By that he evokes trust not only in himself but demonstrates to have trust in the team
members by involving them in his vision. Findings of Edmondson’s (2003) on how team
leaders facilitate learning confirm this argumentation. First, leaders create motivating effort
through a compelling rationale, and second they create psychological safety. In this study
charismatic leadership behavior appeared to have the strongest relationship to team learning
of all investigated leadership behaviors. However, we need to be a cautious with respect to
our conclusions about charismatic leadership behavior. The reliability of our measurement
scale of charismatic leadership is low (cronbach’s alpha = .55). A widely advocated level of
adequacy for Cronbach’s alpha is .70 (Nunally, 1978). For a two-item scale, however, there
seems to be disagreement in the field as to the best indicator of scale reliability (Hulin,
Netemeyer & Cudeck, 2001). Coefficients of homogeneity for any test or scale must be
evaluated against the purpose of the scale, the construct being estimated, and the number of
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
13
items of the test. Items reflecting different dimensions of behavior are correlated but not
redundant. Alpha can be to high if it reflects only a large number of items or extreme wording
redundancy among items. Further research with a charismatic leadership scale, reflecting a
better representation of the dimensions, may result in a more reliable scale that is needed to
draw more certain conclusions. This is also applicable for the other leadership behaviors, that
are investigated with two-item scales in this study.
Team stability explains less variance in team learning (2%) compared to person-
focused leadership (17%). This might be explained by the mixed findings about this
relationship (Edmondson, Bohmer & Pisano, 2001). On one hand a new new or a leaving
team members may disrupt the learning routine in a team, on the other hand such events may
evoke new or increased learning routines. Moreover, the timing of such events might be of
influence as well. A new team member joining a team after building a strong sense of
psychological safety among the old members, might have a more deteriorating influence on
learning than at an earlier point in time. Furthermore, the behavior of the project manager
might have a moderating effect on the relationship between team stability en learning. For
example coaching leadership might help to quickly build a new sense of psychological safety
(Edmondson, Bohmer & Pisano, 2001). More research is needed to shed light on this
relationship.
Task interdependence also explains less variance in team learning (2%) compared to
person-focused leadership (17%). And in a mixed set of all antecedents in this study, task
interdependence even disappears as significantly related antecedent of learning. One
explanation might be that putting people together on a task, only helps a little because it will
increase communication to coordinate. But it might just be not enough to make a substantial
difference in team learning. The behavior of the project manager might have a moderating
effect by involving team members jointly in decision making. A project manager who
challenges his team to produce a collectively supported decision, might also promote team
learning. Another explanation might be that task interdependence in project teams is always
high, which makes it a variable that is not easy to intervene on. Our sample indeed
demonstrated a relative high extent of task interdependence with little variation (M = 4.54,
SD = .65). Further research among project teams in other settings are needed to get more
insights in this relationship.
The current study has several strengths and limitations. A strength of this research lies
in the fact that it was done amongst project teams from different kinds of organizations, which
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
14
means that the findings can probably be generalized to several work settings. However, some
limitations can be outlined as well.
A first limitation lies in the cross-sectional nature of this study. This design does not
allow for testing of directionality of the results. In other words, reverse causality (e.g., team
learning behavior increasing team stability) can not be ruled out based on these data and the
causal ordering should be tested. In order to test for directionality, longitudinal and
experimental research will be necessary.
Second, this study provides a limited exploration of factors that project managers and
their teams can influence in their efforts to promote team learning. Other possible antecedents
of team learning could be collocation of the team members, the presence of a facilitator to the
learning process, and meeting frequency of the team. Future research should assess the
relationship with those and other antecedents.
Third, the measurement instruments of the several leadership behaviors all consisted
of only two-item scales, selected from a larger collection of comparable items designed to
measure a common domain of leadership behavior. A problem of these two-item scales is that
they are not sufficient to represent the larger domain. If the item correlations and variances
are heterogeneous in the domain (as is likely the case with the various leadership behaviors), a
sample of only two items may not detect the heterogeneity. The more desirable approach is to
compute alpha on a set of 10 to 15 items, because a scale of this size is more likely to
incorporate the variability in the domain (Hulin, Netemeyer and Cudeck, 2001).
Finally, all the variables were measured and analysed at the individual level. This
means that the outcomes reflect to what extent individually perceived leadership behavior,
team stability and task interdependence relate to the individually perceived extent of team
learning behaviors. Although the analysis of the individual perspective as regards the
antecedents of team learning has led to some interesting results, using a multi-level approach
will provide additional insights about team-level perspectives.
Some additional avenues for future research can be outlined. It would be interesting to
study whether and how project managers’ leadership behavior can moderate the relationship
of team stability and task interdependence. Theoretically this would be interesting to know
whether it is possible to moderate these relationships. From a practical perspective, such
knowledge could be useful for project managers themselves. It would help them to tailor their
leadership behavior to the team stability and task interdependence at hand. Furthermore it
would be interesting to examine if the leadership behaviors that promote team learning vary
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
15
over time, depending on the project phase. Again this would help project managers to tailor
their behavior to the situation at hand.
Findings of this study contribute from a practitioner as well as an academic
perspective. From a more academic perspective: the ability to learn and the ability to share
what has been learned is mentioned as one of the five major directions for future research in
project management by the Rethinking Project Management network (Winter et. al., 2006).
Findings of this study shed some light on antecedents of team learning in project teams that
can be influenced by the team itself. From a practitioners perspective project oriented
organizations can benefit from the results of this study by helping their project managers how
to increase the learning ability of their teams by using our findings about leadership
behaviour, team stability and task interdependence.
Acknowledgement
The research described in this paper was made possible by a grant from the Project Management Institute Inc, USA. We are grateful for the support given to us by the representatives of PMI®
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
16
References Alvesson, M. (2004). Knowledge Work and Knowledge-Intensive Firms. Oxford University
Press, Oxford, UK.
Akgün, A.E. and Lynn, G.S. (2002). Antecedents and consequences of team stability on new
product development performance, Journal of Engineering and Technology
Management, 19, 263–286.
Aragon S.J., & Gesell S.B. (2003). A Patient Satisfaction Theory and its Robustness Across
Gender in Emergency Departments: A Multi-group Structural Equation Modeling
Investigation. American Journal of Medical Quality, 18(6), 229-241.
Arbuckle, J.L. (2006). Amos 7.0 User's Guide. Chicago: SPSS Inc.
Argote, L., Gruenfeld, D., & Naguien, C. (2001). Group learning in organisations. In: M.E.
Turner (Ed.), Groups at work: Advances in theory and research [pp. 369-411].
Hillsdale, NJ: Erlbaum.
Banker, R.D., Field, J.M., Schroeder, R.D. & Sinha, K.K. (1996). Impact of Work Teams on
Manufacturing Performance: A Longitudinal Field Study. Academy of Management
Journal, 39(4), 867-890.
Bass, B.M. (1985). Leadership and performance beyond expectations. New York: The Free
Press.
Beers, P.J. (2005). Negotiating Common Ground: Tools for multidisciplinary teams. Heerlen:
Unpublished PhD-thesis, Open University of the Netherlands.
Browne, M.W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In: K.A.
Bollen & J.S. Long (Eds.), Testing structural equation models [pp. 445-455].
Newbury Park, CA: Sage.
Burke, C.S, Stagl, K.C. Klein, C., Gooewin, G.F., Salas, E., & Halpin, S.M. (2006). What
type of leadership behaviors are functional in teams? A meta-analysis. Leadership
Quarterly, 17, 288-307.
Campion, M.A., Medsker, G.J. & Higgs, A.C. (1993). Relations between work group
characteristics and effectiveness: Implications for designing effective work groups.
Personnel Psychology, (46)4, pp. 823.
Crawford, J. L., and Haaland, G.A. (1972). Predecisional information-seeking and subsequent
conformity in the social influence process. Journal of Personality and Social
Psychology, 23, 112-119.
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
17
De Jong, R.D. and Carpay, M. (1991). Leadership style and group member’s mood, job
satisfaction and effectiveness. Paper presented at ‘the 5th European Congress on
Psychology of Work and Organization’. Rouan, France.
Den Hartog, D. Van Muijen, J. and Koopman, P. (1994). Transactioneel versus
transformationeel leiderschap. Een analyse van de MLQ in de Nederslandse situatie.
Gedrag en Organisatie, 7(3), 155-166.
Drucker, P. F. (2003). The New Realities. USA: Transaction Publishers U.S.
Edmondson, A. (1996). Learning from mistakes is easier said than done: Group and
organizational influences on the detection and correction of human error. Journal of
Applied Behavioral Science, 32, 5-32.
Edmondson, A. (1999). Psychological safety and learning behavior in work teams.
Administrative Science Quarterly, 44, 350-383.
Edmondson, A.C., Bohmer, R. and Pisano, G.P. (2001). Disrupted Routines: Team Learning
and New Technology Adaptation. Administrative Science Quarterly, 46, 685-716.
Edmondson, A. (2002). The local and variegated nature of learning in organizations.
Organization Science, 13(2), 128-146.
Edmondson, A. (2003). Speaking up in the Operating Room: How Team Leaders Promote
Learning in Interdisciplinary Action Teams. Journal of Management Studies, 40(6),
1419-1452.
Edmondson, A.C., Dillon. J.R., & Roloff, K. (2007). Three Perspectives on Team Learning:
Outcome Improvement, Task Mastery, and Group Process. In: J.P. Walsh & A.P.
Brief (Eds.), The Academy of Management Annals [pp. 269-314]. Hillsdale:
Psychology Press.
Ellis, A.P.J., Hollenbeck, J.R., Ilgen, D.R., Porter, C.O.L.H., West, B.J. & Moon, H. (2003).
Team learning: Collectively connecting the dots. Journal of applied Psychology, 88, 5,
821–835.
Fleishman, E.A., Mumford, M.D., Zaccaro, S.J., Levin, K.Y., Korotkin, A. L., & Hein, M.B.
(1991). Toxonomic efforts in the description of leader behavior. A synthesis and
functional interpretation. Leadership Quarterly, 4, 245-287.
Gibson, C.B., & Vermeulen, F. (2003). A healthy divide: Subgroups as a stimulus for team
learning behaviour. Administrative Science Quarterly, 48, 202-239.
Glassop, L.I. (2002). The organizational benefits of teams. Human Relations, 55(2), 225-249.
Hackman, J.R. (1987). The design of work teams. In: J. Lorsch (Ed.), Handbook of
organizational behaviour. Englewood Cliffs, NJ: Prentice-Hall.
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
18
Hoyle, R.H. (1995). Structural equation modeling: Concepts, issues, and applications.
Thousand Oaks, CA: Sage.
House, R.J. and Mitchell, T.R. (1974). Path-goal theory of leadership: some post hoc and a
priori tests. In: J.G. Hunt & L.L. Larson (eds.). Contingency approaches to
leadership. Carbondale: Southern Illinois University Press.
Hulin, C. , Netemeyer, R., and Cudeck, R. (2001). Can a Reliability Coefficient Be Too
High?, Journal of Consumer Psychology, Vol. 10, No. 1/2, 55-58
Johnson, D.W. (1973). Communication in conflict situations: A critical review of the
research. International Journal of Group Tensions, 3, 46-67.
Judge, T.A., Piccolo, R.F. and Ilies, R. (2004). The forgotten ones? The validity of
consideration and initiating structure in leadership research, Journal of Applied
Psychology, 89, 36-51.
Kasl, E., Marsick, V., & Dechant, K. (1997). Teams as Learners: A Research-Based Model on
Team Learning. Journal of Applied Behavioral Science, 33, 227-246.
Le Blanc, P.M. (1994). De steun van de leiding. Amsterdam: Thesis Publishers.
Marsh, H. W., Balla, J. R., & Hau, K. T. (1996). An evaluation of incremental fit indexes: A
clarification of mathematical and empirical properties. In G. A. Marcoulides & R. E.
Schumacker (Eds.), Advanced structural equation modeling techniques. (pp.315-353).
Mahwah , NJ : Lawrence Erlbaum.
Moreland, R.L. and Levine J.M. (1989). Newcomers and oldtimers in small groups. In: P.B.
Paulus, Editor, Psychology of Group Influence, Lawrence Erlbaum, NJ (1989), pp.
143–186.
Mulder, M., De Jong, R.D., Koppelaar, L. & Verhage, J. (1986). Power, situation and
leaders’s effectiveness: an organizational field study. Journal of Applied Psychology,
71, 566-570.
Mulder, M. Ritsema van Eck, J.R. & De Jong, R.D. (1971). An organization in crisis and non-
crisis situations. Human Relations, 24, 19-41.
Nunnally, J. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill.
Pfeffer, J. (1994). Competitive advantage through people; unleashing the power of the
workforce. Boston: Harvard Business School Press.
Popper, M. and Lipshitz, R. (1992). Coaching on leadership. Leadership and Organizational
Development Journal, 13, 7, 15-18.
Saavedra, R., Earley, P.C., & Van Dyne, L. (1993). Complex interdependence in task-
performaning groups. Journal of Applied Psychology, 1, 61-72.
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
19
Savelsbergh, C., Van der Heijden, B.I.J.M., & Poell, R.F. (2007). “The good, the bad and the
ugly”: Which team competences mediate the relationship between team learning and
performance within project teams? In: D. Jepson & D. Witney (Eds.), Proceedings of
the University Forum for HRD international research conference held in Oxford, UK,
27-29 June. Oxford: Oxford Brookes University.
Savelsbergh, C., B. van der Heijden and R. Poell (2008), “The development and empirical
validation of a Multi-dimensional measuring instrument for team learning behaviors”,
working paper submitted for publication in Small Group Research.
Schippers, M.C., Den Hartog, D.N., Koopman, P.L., & Wienk, J.A. (2003). Diversity and
team outcomes: The moderating effects of outcome interdependence and group
longevity and the mediating effect of reflexivity. Journal of Organizational Behavior,
24(6), 779-802.
Stewart, G.L. and Manz, C.C. (1995) Leadership, membership and organization,
Psychological Bulletin, 47, 1-14.
Starbuck, W. H. (1992) ‘Learning by knowledge intensive firms’, Journal of Management
Studies, (29)6, 713-740.
Stogdill, R.M. (1963). Manual for the Leader Behavior Description Questionnaire, Form XII.
Columbus: Bureau of Business Research, Ohio State University.
Stogdill, R.M. (1950). Leadership, membership and organization, Psychological Bulletin,
47,1-14.
Stoker, J.I. (1998). Leiding geven aan zelfstandige taakgroepen. Unpublished Ph.D.
dissertation, University of Twente, Enschede, Netherlands.
Van den Bossche, P., Gijselaers, W., Segers, M., & Kirschner, P.A. (2006). Social and
Cognitive Factors Driving Teamwork in Collaborative Learning Environments: Team
Learning Beliefs and Behaviors. Small Group Research, 37, 490-521.
Van Dyck, C. (2000). Putting Errors to Good Use: Error Management Culture in
Organizations. Kurt Lewin Institute dissertation series. University of Amsterdam.
Van Dyck, C., Frese, M., Baer, M., & Sonnentag, S. (2005). Organizational error
management culture and its impact on performance: A two-study replication. Journal
of Applied Psychology, 90(6), 1228-1240.
Van de Vegt, G., & Bunderson, S. (2005). Learning and performance in multidisciplinary
teams: The importance of collective team identification. Academy of Management
Journal, 48(3), 532-547.
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
20
Van Woerkom, M. (2003). Critical reflection at work: Bridging individual and
organizational learning. Unpublished Ph.D. dissertation, University of Twente,
Enschede, Netherlands.
Wageman, R. (1995). Interdependence and group effectiveness. Administrative Science
Quarterly, 40, 145-180.
Winter, M., C. Smith, P. Morris and S. Cicmil, 2006, “Directions for future research: the main
findings of a UK Government funded research network”, International Journal of
Project Management, 24, 638-649.
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
21
Table 1: Definitions of the dimensions of Team Learning Behaviors (Savelsbergh, 2008).
New definitions of
Team Learning Behaviors
Item Example and Number of
Items
1. Co-construction of meaning: mutual conversational actions of team
members by refining, building on, or modifying the original offered meaning in some
way to come to ‘new’ meanings in the collaborative work that were not previously
available to the team (based on ‘co-construction of meaning’ of Van den Bossche,
2006).
Information of team members is
completed with information of other
team members.
2. Exploring different perspectives: conversational actions of team members
to explore, share knowledge, opinions and different perspectives (based on
‘construction of meaning’ of Van den Bossche, 2006).
If a team member gives his/her
opinion he/she subsequently asks for
the opinion of the others.
3. Error analysis: discussing and analyzing errors collectively to prevent them
(based on Van Dyck, 2000).
After making a mistake, we try to
analyze what caused it.
4. Error communication: sharing errors collectively to prevent them (based on
Van Dyck, 2000).
Team members communicate their
mistakes, to prevent that others
make the same mistake.
5. Reflection on processes: collectively discuss the team goals, assumptions,
working methods and strategies, checking: is the team doing the right things and
doing things right (based on Reflexivity on processes, Schippers et al., 2003)
We often discuss our team's work
methods.
6. Reflection on outcomes: collectively look back or ahead on experiences and
actions (for example by feedback or communicated errors) to evaluate and learn from
them (based on Reflexivity by evaluating/learning, Schippers et al., 2003).
In our team we check what we can
learn from accomplishments.
7. Feedback seeking behavior: seeking feedback internally among team
members and externally from outside the team in order to reflect (based on Schippers
et al., 2003).
‘We analyze our performance in
accordance with other teams.
8. Experimenting: collectively doing things differently than before and
measuring differences in outcome (based on Van Woerkom, 2003).
We experiment collectively with
other working methods.
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
22
Table 2: Means, standard deviation, reliability coefficients (Crohnbach’s alpha; in italics on the main diagonal), and correlations between the model variables, N = 274. items M SD 1 2 3 4 5 6 7 8 1 Initiating Structure 2 3.01 1.00 0.65 2 Consideration 2 4.33 0.67 0,08 0.70 3 Coaching Leadership 2 3.28 0.98 0,24** 0,44** 0.79 4 Consultative Leadership 2 4.24 0.70 0,02 0,42** 0,40** 0.60 5 Charismatic Leadership 2 3.75 0.74 0,39** 0,29** 0,52** 0,21** .55 6 Team Stability 1 1.39 0.61 -0,11 0,04 -0,01 0,05 -0,07 / 7 Task Interdependence 3 4.54 0.65 0.05 0.18** 0.11 0.16** 0.12* 0.07 0.83 8 Team Learning behavior 28 3.49 0.61 0.10 0.27** 0.34** 0.27** 0.18** 0.15* 0.15* 0.94 ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed)
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
23
Table 3: Significance of relationships of a single antecedent (Initiating Structure, Consideration, Coaching, Consultative & Charismatic team leadership behaviors, Team stability and Task interdependence (independent variables) to team learning behaviors (dependent variable). Beta P Explained
Variance of Team Learning
Initiating Structure - Not significant - Consideration .282 .000 8% Coaching .171 .009 3% Consultative .286 .000 8% Charismatic .362 .000 13% Team Stability .148 .022 2% Task Interdependence .133 .039 2%
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
24
Figure 1: Relationships Person-focused Team leadership behaviors and Team learning behaviors Based on SEM (stand. estimates)(χ2 = 596.500 ; df = 403; χ2 / df =1.48; p = .000; NFI= .867; CFI = .952.; RMSEA = .042; PCLOSE = .973)
,86
Reflection onprocesses
,79
Feedbackseeking
,80
Reflection onoutcomes
,45
Experimentation
,78
Error communication
,87
Error analysis
,83 Exploring
& dif perspec
tives
,57Co Con structionof meaning
,17Team Learning
Behavior
e29
e30
e31
e32
e33
e34
e35
e36
,88
,93
,91
,67
,89
,89
,93
CoachingleadSUM
ConsultativeLeadSUM
ConsiderationLeaderSUM
CharismaticLeadSUM
,14
,28
Dist1
,76
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
25
Figure 2: Relationships of the Combined set of individually significantly related antecedents to Team learning behaviors Based on SEM (stand. estimates)(χ2 = 728.453 ; df = 466; χ2 / df =1.56; p = .000; NFI= .843; CFI = .935.; RMSEA = .045; PCLOSE = .883)
,86
Reflection onprocesses
,78
Feedbackseeking
,79
Reflection onoutcomes
,44
Experimentation
,78
Error communication
,86
Error analysis
,83Exploring
& dif perspec
tives
,57Co Con structionof mea
ning
,18Team Learning
Behavior
e29
e30
e31
e32
e33
e34
e35
e36
,88
,93
,91
,67
,89
,89
,93
CoachingleadSUM
ConsultativeLeadSUM
TaskinterdependenceSUM
ConsiderationLeaderSUM
CharismaticLeadSUM
TeamStability
,28
,14
Dist1
,76
© 2008, Chantal Savelsbergh, Peter Storm, Ben Kuipers Sponsored by PMI-America
26
Appendix 1:
,86
Reflection onprocesses
,78
Feedbackseeking
,80
Reflection onoutcomes
,45
Experimentation
,78
Errorcommunication
,87
Erroranalysis
Second Order Model ofTEAM LEARNING BEHAVIORS
Chisquare = 470,787; Df = 295; P = ,000GFI = ,876NFI = ,885CFI = ,953
RMSEA = ,047; PCLOSE = ,749
,83Exploring
& difperspec
tives
,57Co Constructionof mea
ning
Team LearningBehavior
e29
e30
e31
e32
e33
e34
e35
e36
,88
,93
,91
,67
,88
,89
,93
,75