| Abstract: | Despite the common use of consensus amongst experts as a “surrogate for truth” in experimental studies of decision-making, there has been little discussion as to its appropriateness. A new measure of consensus for dichotomous predictions, which takes into account expert' confidence in the correctness of their decisions, is proposed and its relationship with accuracy is experimentally evaluated. The experiment also extends Ashton's (The Accounting Review, April 1985, pp. 173–185) empirical work on the conventional consensus measure by checking the robustness of Ashton's results to a different experimental setting (U.K. trade credit specialists) and to increases in the cues made available to experimental participants. The paper concludes that the new measure of consensus complements the conventional measure, in that it provides information over and above that provided by the conventional consensus measure. The paper also concludes that the results noted by Ashton for the U.S. also hold for the U.K. and are robust to increases in the cue set made available to participants. |
