Analyzing the Effect of Team Structure on Team Performance: An Experimental and Computational Approach

This paper investigates how team structure impacts team information processing and how this affects team resilience to change. Teams in three different subgrouping structures: homogeneous, heterogeneous, or no-clustering, were instructed to make decisions in which they had to evaluate different companies and pick the best one. Each member was told to evaluate the companies in a different way, creating teams with diverse perspectives. Partway through the experiment the problem evaluation criteria changed, but teams were not informed of the change, only whether their choice was correct. Teams with homogeneous-clustering were less capable than the other two types of teams in making use of multiple preferences and in dealing with changes. A similar effect was also found in computational simulations built from a PDP model. We suggest that heterogeneous-clustering can weaken members’ ownership and confirmation biases while no-clustering ensures a free flow of information, with both able to enhance team performance.