Estimating prospect theory's decision weights with stochastic dominance: The small probability case

When one prospect is certain and the other uncertain, Cumulative Prospect Theory employs the certainty equivalent methodology to estimate Decision Weights (DW). However, DW may be different with two uncertain prospects. In this study, we neutralize the "certainty effect"and propose Stochastic Dominance (SD) to estimate DW for the first time with small probabilities, which is the raison d'être of the employment of DW. Using SD we provide ranges, rather than point estimates, of DW parameters that are consistent with all possible S-shape value functions. Comparing CE and SD implied DW, we find that DW are situation dependent: DW derived with one certain prospect are much different than those derived with two uncertain prospects.