ABSTRACT
Decision-making groups must often choose froma set of specified alternatives (e.g., guilty and not guilty for a jury, a set of applicants for a selection committee), and this choice is typically preceded by discussion of the merits of each alternative. One goal of discussion is to achieve a consensus among the group’s members. Another goal is to pool members’ expertise and knowledge. In principle, pooling information permits a group decision that is more informed than the decisions of members acting individually. In particular, discussion can perform a corrective function when members individually have incomplete and biased information but collectively can piece together an unbiased picture of the relative merits of the decision alternatives. Notwithstanding the potential of discussion to serve such a corrective function, group discussion may often fall short of its potential. Discussion is rarely a systematic and balanced exploration of the relevant issues. On the contrary, it is often thematic and consensus confirming; that is, discussion tends to focus on particular issues and to support an existing or emergent consensus (cf. Fisher, 1980). Such patterns may counter effective pooling of information and may perpetuate biases that members bring to the group.
