The Effect of Task-Relevant Diversity on Small Group Performance
Researchers have examined different types of diversity (e.g., demographic diversity, cognitive diversity, etc.). In this paper, I examined how task-relevant diversity (i.e., the diversity of statistically extracted attributes of individuals’ responses to the tasks) and group size influence artificial group performance. I generated homogeneous, random, and diverse artificial groups based on the task-relevant attributes extracted from individuals’ responses to the tasks from the NFL predictions dataset and the AP Math dataset. The group performance of these three groups and the best member of the diverse groups were compared. The study of the NFL predictions and the AP Math dataset showed that task-relevant diversity had a negative impact on group performance, and small homogeneous groups could compete with large homogeneous groups. Further simulations and manipulation of the distribution of the performance of the original sample showed that task-relevant diversity had a positive impact on group performance when the majority of the original sample performed poorly or when the original sample performance was normally distributed. This study also suggested that group size did not necessarily have a positive effect on group performance, in contrary to previous research.
Zhang, Yizhi, "The Effect of Task-Relevant Diversity on Small Group Performance" (2021). ETD Collection for Fordham University. AAI28714828.