G-I transfer and coordination gains: Group learning in discretionary group tasks

点击翻译按钮获取中文摘要

The accuracy of group judgments often exceeds the accuracy of the average of the group members` individual judgments. In the first phase of this project, we successfully demonstrated that this process gain is not, as currently assumed in the literature, due to coordination gains (i.e., a stronger weighting of more accurate individual proposals), but rather due to G-I transfer: As a consequence of collaborative work, particularly during the first group interaction, weaker group members experience a socially induced increase in their individual accuracy, and this increase subsequently also benefits group performance. The first set of experiments of the renewal proposal submitted here aims at generalizing this G-I transfer to group forecasts (a judgment type with particular practical relevance). Furthermore, we test to what extent communication is necessary for this transfer to occur. In addition, we want to provide a more detailed temporal and content-wise analysis of this transfer by distinguishing between mapping errors and metric errors in group judgment. In the second set of experiments of this proposal, we want to systematically test whether coordination gains in group judgment depend on the presence of a population bias in the judgment task, that is, whether these gains typically occur if group members share a tendency to uniformly under- or overestimate the true values. In addition, we want to specify the processes that enable group to exhibit such coordination gains in these types of tasks.