Disaggregated Probabilistic Accounting Information: The Effect of Sequential Events on Expected Value Maximization Decisions

Abstract

In a previous study reported in this Journal, Ronen [1971] investigated user reaction to disaggregated probabilistic information given about sequential events. The purpose of his study was to determine the effect of providing probabilistic information on sequential events on decision makers' use of expected value maximization rules. In a two-act decision where each act had two events occurring in sequence, Ronen found that when the joint probabilities of success of the two acts were equal (i.e., overall expected values of the alternatives were equal), subjects preferred the act where the probability of success associated with the first event (initial probability) was higher. (This choice will hereafter be termed the sequence effect.) Subjects in his study received both aggregated and disaggregated probabilistic information analogous to that shown in table 1. Subjects preferred Act 1 in this example since p(E,12A1) > p(E12A,2). Ronen also made a limited investigation of two-act decisions where the joint probabilities were unequal. Subjects showed a preference for the act with the higher initial probability but the lower overall expected value. This preference was found when the difference between the expected values of the two acts was less than a specific level.