Bayesian model comparison favors quantum over standard decision theory account of dynamic inconsistency.

Abstract

Many paradoxical findings in decision-making that have resisted explanations by standard decision theories have accumulated over the past 50 years. Recent advances based on quantum probability theory have successfully accounted for many of these puzzling findings. Critics, however, claim that quantum probability theory is less constrained than standard probability theory, and hence quantum models only fit better because they are more complex than standard decision models. In this article, for the first time, a Bayesian method was used to quantitatively compare the 2 types of decision models, which is a method that evaluates models with respect to accuracy, parsimony, and robustness. A large experiment was used to compare the best-known models of each type, matching in their numbers of parameters, but possibly differing in the complexity of their functional forms. Surprisingly, the Bayesian model comparison overwhelmingly favored the quantum model, indicating that its success is due to its robust ability to make accurate predictions rather than accidental fits afforded by increased complexity.