Before proceeding, I should make clear that, although this chapter is fundamentally about the composition of task performing groups, there are many compositional issues that are not addressed. Group size, staffing levels, and the average standing of group members on particular variables are three examples. I ignore these not because they are uninteresting or have no bearing at all on group performance, but because they seem less likely to shed light on the question of synergy in groups. A case in point is the average cognitive ability of group members. There is good evidence that a group’s performance is related to the mean cognitive ability of its members; the higher their average cognitive ability score, the better their collective performance. This effect holds across a wide range of tasks and is mediated by such factors as group learning and the development of

shared mental models (e.g., Day et al., 2004, 2005; Devine & Philips; 2001; Edwards, Day, Arthur, & Bell, 2006; Ellis et al., 2003). However, effects like this reveal relatively little about what might be occurring during group interaction. As argued in Chapter 1, synergy is an emergent phenomenon rooted in the interactions of group members as they perform their task together. To the extent that synergy is manifest at all in a particular group’s performance, it is more apt to be conditioned on the way variables are distributed across members than on the central tendency of those variables. Measures of dispersion and variability are inherently relational and so quite naturally lead us to focus on the group’s internal dynamics and interaction processes. As such, the study of diversity, more than the study of average member characteristics, has the potential to illuminate factors that either promote or inhibit the emergence of synergy within groups.