Assessment of Group Preferences and Group Uncertainty for Decision Making.

摘要

Often groups rather than individuals function as decision makers. To use decision analysis to aid such decision makers, group preferences and opinions must be quantified as utilities and probabilities. This paper reviews the procedures by which these group utilities and probabilities can be determined, along with the relative merits of individual versus group judgments of these quantities. While group judgments are generally superior to individual judgments, no entirely satisfactory method for determining either group utilities or probabilities exists. Using ordinal preferences rather than cardinal utilities, majority rule will lead to a satisfactory group preference function in certain restricted situations. Both mathematical aggregation rules and behavioral approaches have been suggested for obtaining group probabilities. The mathematical approaches range from the simple to complex. The theoretically elegant methods, however, generally suffer in practice from the impossibility of determining some of the inputs necessary for the models. The behavioral approaches attempt to reduce disagreement by communication and interaction among the group members. The most successful methods depend on structural communication to allow the facilitation of group judgments while avoiding many of the detrimental influences that have been identified in social psychological research.