How to Improve Decision Making in Small Groups

Abstract

Decision-making groups are often biased in favor of shared information (sharedness bias) and in favor of its members' initial preferences (preference bias). The present experiment aimed at analyzing both biases at the group level (communication of information and preferences) and at the individual level (evaluation of information) simultaneously. Two interventions were evaluated, each focusing on one of the two biases and illustrating it with a group exercise. The interventions enhanced the amount of discussed information and reduced the preference bias but had no effects on decision quality. Dissent (diversity in members' initial preferences) enhanced the preference bias in information exchange but reduced both biases in information evaluation and improved decision quality when no intervention was applied. Decision quality correlated with individual-level processes but not with group-level processes.