Hierarchical Bayesian parameter estimation for cumulative prospect theory ?

摘要

Nilsson, Rieskamp, and Wagenmakers (2011) implemented a hierarchical Bayesian estimation pro-cedure for cumulative prospect theory (CPT; Tversky and Kahneman (1992)). Nilsson et al. used a simplified version of CPT that holds for choice options with mixed outcomes, that is, one positive and one negative payoff. However, for choice options with only gains or only losses, the model specification does not hold. Here we provide a corrected model specification of CPT, one that also holds for options with only gains or only losses. We show that the corrected version does not change the qualitative results reported in Nilsson et al. (c) 2020 Elsevier Inc. All rights reserved.