The Danger of Disappearing Postdecision Surprise: Comment on Harrison and March, "Decision Making and Postdecision Surprises"

Abstract

Bernard Goitein In the March 1984 issue of ASQ, Harrison and March (Decision Making and Postdecision Surprises) mathematically derive the existence of a structural bias toward postdecision disappointment. They do so by building on research in behavioral decision theory. This research has shown that intuitive predictions of outcomes are nonregressive, i.e., they do not obey the statistical principle of regression toward the mean (Kahneman and Tversky, 1973). Expectations therefore tend to be more extreme than actual outcomes of decisions, producing a preponderance of disappointments among postdecision surprises (Harrison and March, 1984).1 If the structural bias toward disappointment is straightforwardly translated into the distribution of subjective surprise, then decisions, even good decisions, are likely to be disappointing, because their outcomes fail to meet the decision maker's expectations. If, however, people have developed effective mechanisms for coping with the structural bias, then Harrison and March's finding may be only an interesting mathematical demonstration.