Decision Making under Strict Uncertainty: An Experimental Test of Competitive Criteria

Abstract

In decisions under strict uncertainty individual decision makers have to choose one of a finite number of alternatives with complete information about their outcomes but in the absence of any information about the probabilities of the various states of nature. We report two experiments with multiple decision tasks in which subjects were instructed to rank the alternatives (Experiment 1), or choose the best one (Experiment 2). We test the major decision rules (criteria) which have been proposed for this case in the normative literature with special attention on Hurwicz′s "optimism-pessimism" criterion which is the only one that can accommodate individual differences. Although none of the criteria can account for the results of the majority of the subjects, choices made by a sizable minority of the subjects′ are consistent with Hurwicz′s model. There is no evidence for a systematic relationship between the model′s single "optimism-pessimism" parameter (inferred form choices) and, independently derived, individual measures of optimism (obtained from responses to personality scales). We conclude with a discussion of the difficulty of experimentally implementing strict uncertainty.