Weak Experimental Verification of the Expected Utility Hypothesis .

The five experiments described in this paper are the modest final product of an initially ambitious attempt to verify a theory of choice under uncertainty. Yaari [5] in 1965 reported the results of some experiments he had performed in which he attempted to show that the acceptance of unfair gambles is better explained by the assumption that gamblers are optimistic about the occurrence of low probability events, than by the assumption of increasing marginal utility of wealth as postulated by Friedman and Savage [1] and as demonstrated experimentally by Mosteller and Nogee [2]. The hypotheses I attempted to test in these five experiments all rest on the assumption that a subject faced with a choice under uncertainty will calculate the expected utility of each alternative, in the sense of yon Neumann and Morgenstern [3], and will choose that alternative for which the expectation is greatest. In the first and second experiments I was testing a theory having to do with the formation, by the subject, of his beliefs as to the proper probabilities to use in the calculation. In the third I was interested in verifying Yaari's experimental findings, and in the fourth and fifth experiments I was trying merely to obtain very weak evidence to support the hypothesis that decisions are made under uncertainty so as to maximize expected utility. [ABSTRACT FROM AUTHOR]