Hierarchical Bayesian Estimation of Quantum Decision Model Parameters

Abstract

Quantum decision models have been recently proposed to account for findings that have resisted explanation by traditional decision theories. This paper compares quantum versus Markov models of decision making for explaining a puzzling empirical finding from human decision making called dynamic inconsistency – that is the failure of decision makers to carry out their planned decisions. A large data set that empirically investigated dynamic inconsistency was used to quantitatively evaluate the quantum and Markov models. In this application, the quantum model reduces to the Markov model when one of the parameters is set to zero. The parameters of the quantum model were estimated using Hierarchical Bayesian estimation. The distribution of the key quantum parameter was clearly located in the quantum regime and far below zero as predicted by the Markov model. These results provide further support for quantum models as compared to the traditional models of decision making.