The Analysis of Activity and Accuracy in Group-Based Forecasting
This study investigates whether group size and different types of priming have an effect on activity and accuracy in group-based forecasting, utilizing the forecasting results from a 4-year geopolitical forecasting tournament. Participants were randomly divided into two team sizes (5 vs. 15), and three types of priming of background of group members (control, heterogeneous interests, and homogeneous interests. We also simulate two types of nominal teams from individuals and teams of 5, and compare their performance to the real teams’. Results demonstrate (1) that smaller teams are more active than larger teams, (2) that teams who are primed to value diversity of interests and opinions are more active than teams who are primed to value homogeneity of interests, (3) that nominal teams based on individuals are more active than nominal teams based on teams of 5, which are more active than real teams, (4) that nominal teams based on teams of 5 produced the more accurate results than nominal teams based on individuals and real teams.
Zhang, Yizhi, "The Analysis of Activity and Accuracy in Group-Based Forecasting" (2017). ETD Collection for Fordham University. AAI10620585.