Abstract

Summary The failure of earlier studies to find much evidence for differential accuracy may have been due to the methodology employed. Two modifications of the standard methodology were introduced in the present experiments: (a) the judges self-selected the information on which their judgments were based, and (b) the objective value of the information was specified. Groups of judges selected information with and without behavioral context, and received it with and without behavioral context. All groups achieved significantly greater differential accuracy than the control group. Although judges varied widely in their information preferences, the more accurate judges did not choose objectively better information. Behavioral context significantly increased accuracy only when provided during the information selection process. The second experiment replicated the major findings and further clarified the function of behavioral context. The results underline the necessity of distinguishing between objective and subjective information value. Judges are capable of differential accuracy when allowed to choose information of value to them. Behavioral context for the information significantly improves their choice.