The consequences of role-conferred bias and base-rate neglect

This study examines the effect of training to reduce biases and heuristics on the consequence of judgements. We demonstrate that untrained subjects' judgments may systematically yield better consequences than judgments of subjects trained to reduce biases and heuristics. This result implies that educators should use caution when interpreting the findings of biases and heuristics research. In establishing the existence of situations in which biases and heuristics produce better consequences than an unbiased, or nonnative procedure, we open the larger question of the need for an elaboration of the prescriptive uses of the normative procedures.