On The Use of Consensus: Prediction, Attribution and Evaluation

Abstract

Subjects read about a psychology experiment on helping be havior. Consensus information was manipulated by presenting base rates for experimental helping behavior (high, low, or no base rate information) and by altering the described situation such that helping behavior was more or less expected of participants (high, medium, low or no implicit consensus). Base rate information influenced predic tions more than did implicit consensus information. However, implicit consensus but not base rates influenced causal attributions for non- helping behavior and evaluations of non-helpers. Implications of these results for the recent dispute concerning observers' use of consensus information were discussed.