Although diversity has often been proposed to improve group performance, its impact on group cooperation has largely been overlooked. This article uses a theoretical approach to examine the effects of diversity on group tasks in the context of the free-rider problem. We identified 3 kinds of diversity that have been commonly reported to improve a group’s performance: (1) ability, (2) cognitive style, and (3) information sources. Each type of diversity was formalized within an evolutionary simulation based on a foraging metaphor. The simulations were run under different environmental assumptions, covering complexity, costs for cooperation, cue redundancy, and cue compensatory level. Our results indicate that diversity in cognitive style and information sources generally increase cooperation. Both diversity factors also improve group members’ average outcome in non-compensatory environments. However, the outcome results are mixed in compensatory environments. We did not find robust and reliable effects of diversity in ability. Our study provides the first approach to modeling isolated diversity effects on group decision-making in the face of the free-rider problem. It may serve as a theoretical framework for future studies.

Following up on a recent debate, we examined advice taking in dyads compared to individuals in a set of three studies (total N = 303 dyads and 194 individuals). Our first aim was to test the replicability of an important previous finding, namely that dyads heed advice less than individuals because they feel more confident in the accuracy of their initial judgments. Second, we aimed to explain dyads’ behavior based on three premises: first, that dyads understand that the added value of an outside opinion diminishes when the initial pre-advice judgment is made by two judges rather than one judge (given that the dyad members’ opinions are independent of each other); second, that they fail to recognize when the assumption of independence of opinions does not hold; and third, that the resistance to advice commonly observed in individuals persists in groups but is neither aggravated nor ameliorated by the group context. The results of our studies show consistently that previous findings on advice taking in dyads are replicable. They also support our hypothesis that groups exhibit a general tendency to heed advice less than individuals, irrespective of whether the accuracy of their initial judgments warrants this behavior. Finally, based on the three assumptions mentioned above, we were able to make accurate predictions about advice taking in dyads, prompting us to postulate a general model of advice taking in groups of arbitrary size.

In this article, a leadership model is presented, with which to investigate the relationship of trait emotional intelligence (trait EI), leadership self-efficacy and leader's task self-efficacy with collective task efficacy and group performance. The sample was made up of 217 undergraduate students, randomly assigned to work teams of 1 leader and 2 followers that were requested to perform a production task. An adapted version of the Schutte Self-Report Inventory (SSRI; Schutte et al., 1998) was used to measure trait EI. Structural equation modeling was used to test the hypothesized relationships. Results indicated that task self-efficacy was a mediator between leadership self-efficacy and collective task efficacy; the latter, in turn, was the best predictor of group performance. No significant relationship was found between trait EI and collective task efficacy although, unexpectedly, trait EI was positively associated with leadership self-efficacy. Implications of the results are discussed.