Content area
Abstract
The independent effects of presentation order and knowledge of outcomes (e.g., hindsight bias and judgment by outcomes) have been well established in the respective literatures as a significant influence on human judgment. The current study was conducted to test whether the judgment-by-outcomes effect could be explained as a special case of order effect. Typical research in judgment-by-outcomes literature provide subjects outcome information last in the information sequence suggesting a recency effect. This question was examined using the Belief-Adjustment Model proposed by Hogarth and Einhorn (1992) to predict order effects. It incorporates several task characteristics as well as cognitive processes believed to be used by decision makers.
Using 254 college students in a fully randomized 3 x 2 x 2 incomplete factorial design, the effects of outcome information order (i.e., presented first, last, or not at all), outcome sign (i.e., successful or unsuccessful outcome), and decision-process information (i.e., positive-negative and negative-positive order) were examined as to their influence on ratings of the way a leader made decisions, of the decision itself, and of responsibility for the outcomes of that decision.
Results provide partial support for a recency effect of outcome information. An outcome sign-by-order interaction was observed with positive outcome information presented last influencing ratings more than this information presented first. This effect was not found for negative outcome information. Recent outcome information influenced ratings of the recommendation decision more than this information presented first. A marginal effect was found for leader process ratings at p $<$.10. Strong support for the judgment-by-outcomes effect was found with all ratings of leader performance and responsibility significantly influenced by outcome sign. Positive outcome ratings were higher than negative outcome ratings on the performance dimensions. Subjects rated the leader as more responsible for negative outcomes than for positive outcomes. A significant order effect was also observed for process information where recent process information significantly influenced leader performance ratings, but not ratings of responsibility.
