Many theories have been developed to model decision behavior under ambiguity. In this paper we empirically investigate Choquet expected utility, which is based on nonadditive probabilities. We first replicate Ellsberg paradox behavior. Then we elicit nonadditive probabilities (called capacities) for individual decision makers. Those capacities did not have all properties theoretically required by CEU. Finally we found that CEU is not really superior to EU in predicting the participants' decisions.
Dieser Eintrag ist Teil der Universitätsbibliographie.