Content uploaded by Anita Williams Woolley
Author content
All content in this area was uploaded by Anita Williams Woolley on Dec 19, 2015
Content may be subject to copyright.
Cognitive Diversity, Collective Intelligence, and
Learning in Teams
ISHANI AGGARWAL, Tilburg University
ANITA WILLIAMS WOOLLEY, Carnegie Mellon University
CHRISTOPHER F. CHABRIS, Union College
THOMAS W. MALONE, Massachusetts Institute of Technology
1. INTRODUCTION
Work performed in organizations of all types and sizes is increasingly organized through the
mechanism of teams. Indeed, more and more of the most important achievements in our society are
the products of group collaboration. For instance, the proportion of scientific articles and patent
applications produced by groups of collaborators has risen sharply in recent decades, with
collaboratively-produced innovations six times more likely to become “mega-hits” than products of solo
practitioners (Wuchty, Jones, & Uzzi, 2007).
To be successful, these groups must somehow combine the contributions of different members, and a
great deal of attention has recently been focused on how group performance is affected by the diversity
of group members (van Knippenberg & Schippers, 2007; Joshi & Roh, 2009). Much of this work,
however, has focused on various kinds of demographic diversity such as race and gender. In this paper,
we focus on a different kind of diversity: cognitive diversity, or more precisely, diversity in cognitive
style. We examine its effect on two novel measures of group performance: collective intelligence
(Woolley, Chabris, Pentland, Hashmi, & Malone, 2010) and team learning (e.g., Argote & Epple, 1990;
Edmondson et al, 2007). In doing so, we find that (a) moderate levels of cognitive diversity maximize
collective intelligence, (b) collective intelligence is correlated with team learning, and (c) collective
intelligence mediates the relationship between cognitive diversity and team learning.
2. THEORETICAL BACKGROUND
Cognitive styles define ways in which individuals encode, process, and communicate information, and
are related to their functional and educational specializations (Ausburn & Ausburn, 1978;
Kozhevnikov, Kosslyn, & Shephard, 2005). Having the right amount of cognitive style diversity is
important for team performance. Teams with too little cognitive diversity may lack the cognitive
capacity to tackle tasks that require different ways of encoding and processing information, but teams
with too much cognitive diversity may find it difficult to for members with different information
processing styles to communicate effectively. Hence, the possible diversity advantage of having the
increased perspectives and skills within the team are likely to be offset by the excessive coordination
efforts the team has to engage in to benefit from these resources (Aggarwal & Woolley, 2013; Gibson &
Cohen, 2003; Steiner, 1972). Given this trade-off, we hypothesize a curvilinear (inverted U-shaped)
relationship between team cognitive style diversity and collective intelligence, such that the highest
levels of collective intelligence are likely to emerge in teams with a moderate level of cognitive style
diversity.
In addition, we predict that collective intelligence will be strongly associated with learning in teams
(Argote, Gruenfeld, & Naquin, 2001; Argote, 2011). Prior research strongly suggests that collective
Collective Intelligence 2015
1
2 I. Aggarwal, A.W. Woolley C.F. Chabris, T.W. Malone
intelligence in teams is associated with team members’ abilities to perceive subtle nonverbal cues as
well as to exchange information effectively via verbal communication (Engel, Woolley, Jing, Chabris, &
Malone, 2014; Woolley et al., 2010). Additionally, the literature on individual intelligence (“IQ”) shows
that people with more cognitive resources learn faster, both explicitly and implicitly (Chabris, 2007).
These findings suggest that collective intelligence ought to enable not only high performance at any
given point in time, but also improved performance over time, even in tasks that do not require a team
to explicitly learn a new procedure or body of knowledge.
Further, we also investigate the role of collective intelligence as a mechanism through which team
diversity influences team learning. Overall, we anticipate that when the team has the right amount of
diversity—requisite amount of cognitive resources without the coordination difficulties—it will be an
ideal condition for the team’s collective intelligence to emerge, which is then likely to impact the rate
at which a team can learn with experience. That is, any effect of cognitive diversity on learning will be
mediated via the mechanism of collective intelligence.
Hypothesis 1: There will be a curvilinear—inverted U-shaped—relationship between cognitive style
diversity and collective intelligence in teams.
Hypothesis 2: Collective intelligence in teams will be positively correlated with teams’ rate of
learning.
Hypothesis 3: Collective intelligence will mediate the indirect relationship between cognitive style
diversity and team learning.
3. STUDY AND MEASURES
The sample consisted of 337 participants, randomly assigned to 98 teams of two to five participants
each. The minimum-effort tacit coordination game (Van Huyck, Battalio, & Beil, 1990) was used as a
measure of team learning. Games like this one are used to explore the ability of a group of people to
implicitly coordinate their strategy; the game involves multiple rounds of individual decision-making
in which the team gains or loses money as a result of the decisions made by its members, who make
their decisions simultaneously and without communication. This task was conducted in 10 rounds. In
each round, each team member chose a number: 0, 10, 20, 30, or 40. At the end of the round, each
member received points defined by a payoff matrix that took into account the member’s own choice
and the minimum of all member choices on that round. It differed from the standard prisoner’s
dilemma game in that teams are rewarded more for coordinating than for defecting or competing. The
safest choice for individuals is to exert minimal effort (choose 0), which maximizes their minimum
possible payoff. However, in order to maximize earnings, each group member must exert maximal
effort by choosing 40 (Deck & Nikiforakis, 2012), running the risk that someone else will shirk (e.g.,
choose 10 or 0), which could lead to low or negative earnings.
Cognitive styles were captured using the Object-Spatial Imagery and Verbal Questionnaire (OSIVQ)
(Blazhenkova & Kozhevnikov, 2008), which assigns the individual separate scores on the styles of
object visualization, spatial visualization, and verbalization. Cognitive style diversity was analyzed as
the sum of the within-team standard deviations in each cognitive style. Collective intelligence was
measured as the factor capturing the team’s performance across a battery of tasks, as described in
Woolley et al. (2010). Team learning was calculated as the rate of change (or slope) in earnings for each
group across the ten rounds of the game.
Collective Intelligence 2015
Collective Intelligence, Cognitive Diversity, and Learning in Teams 3
4. RESULTS
Hypothesis 1, predicting a curvilinear (inverted U-shaped) relationship between cognitive style
diversity and collective intelligence, was supported. The results of a regression analysis, controlling
for team size and cognitive style means, demonstrated that the quadratic relationship between
cognitive style heterogeneity and collective intelligence was negative (an inverted U-shaped
relationship) and significant: β =
-.91, t = -2.19, p = .03, R2= .40.
Hypothesis 2, predicting a positive
relationship between collective
intelligence and learning, was also
supported (β =.29, t = 2.74, p=.007,
R2= .34, controlling for the
intercept and team size; see
Figure 1). There was no significant
relationship between collective
intelligence and the teams’ choice
at Time 1, which rules out the
alternative explanation that teams
with a higher collective
intelligence had more coordination
at the beginning of the task.
Finally, Hypothesis 3, predicting
collective intelligence as a
mediator of the relationship
between cognitive style diversity
and learning, was supported. There was
no significant main effect of cognitive
style diversity on rate of learning.
Mediation analyses indicated that
there was an indirect relationship between cognitive style diversity and team learning through
collective intelligence. This indirect relationship was evident at high levels of cognitive style diversity
(+1 SD); Ɵ=-1.27. A bootstrap analysis revealed that the 95% bias-corrected confidence interval for the
size of the indirect effect excluded zero (-3.7, -.21), suggesting that high levels of cognitive style
diversity curb team learning indirectly by reducing collective intelligence.
5. CONCLUSION
Teams that have the ability to perform effectively across changing contexts, and align their member
resources into processes that yield consistency in performance, are likely to be more beneficial for
organizations than teams that collapse as soon as there is a change in the environment. This study
Collective Intelligence 2015
Figure 1. The relationship between team collective
intelligence (low versus high) and team learning (round-
by-round earnings in the minimum effort tacit
coordination game).
4 I. Aggarwal, A.W. Woolley C.F. Chabris, T.W. Malone
shows that a moderate amount of cognitive style diversity facilitates the collective intelligence of a
team, which further positively impacts the rate at which teams learn to implicitly coordinate, and that
cognitive style diversity indirectly influences team learning through collective intelligence. Taken
together, our results highlight the importance of collective intelligence as a central construct for
understanding the drivers of team performance.
Collective Intelligence 2015
Collective Intelligence, Cognitive Diversity, and Learning in Teams 5
REFERENCES
Aggarwal, I., & Woolley, A. W. (2013). Do you see what I see? The effect of members’ cognitive styles on team processes and errors in task
execution. Organizational Behavior and Human Decision Processes, 122(1), 92–99.
Argote, L. (2011). Organizational learning research: Past, present and future. Management Learning, 42(4), 439–446.
Argote, L., Gruenfeld, D., & Naquin, C. (2001). Group learning in organizations. In M. E. Turner (Ed.), Groups at work: Theory and research (pp.
369–411).
Ausburn, L. J., & Ausburn, F. B. (1978). Cognitive styles: Some information and implications for instructional design. ECTJ, 26(4), 337–354.
Blazhenkova, O., & Kozhevnikov, M. (2008). The new object-spatial-verbal cognitive style model: Theory and measurement. Applied Cognitive
Psychology, 23(5), 638–663.
Chabris, C.F. (2007). Cognitive and neurobiological mechanisms of the Law of General Intelligence. In M.J. Roberts (Ed.), Integrating the mind:
Domain specific versus domain general processes in higher cognition (pp. 449–491). Hove, UK: Psychology Press.
Deck, C., & Nikiforakis, N. (2012). Perfect and imperfect real-time monitoring in a minimum-effort game. Experimental Economics, 15(1), 71–88.
Engel, D., Woolley, A. W., Jing, L. X., Chabris, C. F., & Malone, T. W. (2014). Reading the Mind in the Eyes or Reading between the Lines?
Theory of Mind Predicts Collective Intelligence Equally Well Online and Face-To-Face. PloS One, 9(12), e115212.
Gibson, C. B., & Cohen, S. G. (2003). Virtual teams that work: Creating conditions for virtual team effectiveness. San Francisco: John Wiley &
Sons.
Ilgen, D. R. (1999). Teams embedded in organizations: Some implications. American Psychologist, 54(2), 129.
Joshi, A., & Roh, H. (2009). The role of context in work team diversity research: A meta-analytic review. Academy of Management Journal, 52,
599-628.
Kozhevnikov, M., Kosslyn, S., & Shephard, J. (2005). Spatial versus object visualizers: A new characterization of visual cognitive style. Memory &
Cognition, 33(4), 710–726.
Steiner, I. D. (1972). Group Processes and Productivity. New York: Academic Press.
Van Huyck, J. B., Battalio, R. C., & Beil, R. O. (1990). Tacit coordination games, strategic uncertainty, and coordination failure. The American
Economic Review, 80(1), 234–248.
van Knippenberg, D., & Schippers, M. C., (2007). Work group diversity. Annual Review of Psychology, 58, 515-541.
Woolley, A. W., Chabris, C. F., Pentland, A., Hashmi, N., & Malone, T. W. (2010). Evidence for a Collective Intelligence Factor in the Performance
of Human Groups. Science, 330(6004), 686–688.
Wuchty, S., Jones, B. F., & Uzzi, B. (2007). The increasing dominance of teams in production of knowledge. Science, 316(5827), 1036–1039.
Collective Intelligence 2015
