中国科学院心理研究所机构知识库

If each member of a group makes less accurate decisions than those of another group, can the former actually make more accurate decisions collectively than the latter? Through four simulation studies, the chapter shows conditions under which such "less-ismore" effect may occur. In each study, a group member adopted either the take-thebest or the minimalist heuristic to make an individual decision, and a simple majority rule was then applied to determine the group decision. Although an individual using take-thebest can generally achieve higher decision accuracy than one using the minimalist, results in Study 1 show that the decision accuracy of a group of take-the-best individuals can be lower than that of a group of minimalist individuals in task environments where the distribution of cue validities is relatively flat. Similar less-is-more effects are found in Studies 2 and 3, where a group of less accurate individuals, due to either their usage of erroneous cue information or cue orders differing from cues' validity order, can outperform another group of more accurate individuals. Finally, the chapter compares the decision accuracy of five-member groups with varying compositions of take-the-best and minimalist members, and found that groups with either one or two take-the-best members can achieve the most robust performance across four task environments. Informational diversity and characteristics of task environments are the main factors underlying the observed less-is-more effects. Therefore, the chapter argues that to understand the rationality of group decision making, these two factors, in addition to the competency of group members, must be taken into consideration. 2013 by Ralph Hertwig and Ulrich Hoffrage. All rights reserved.