Published November 3, 2026 | Version v1

Experiential value neglect in hybrid choice is robust to learning and representational manipulations

  • 1. Laboratoire de Neurosciences Cognitives Computationnelles, Institut National de la Santé et Recherche Médicale, Paris, France
  • 2. Département d'Études Cognitives, École Normale Supérieure, Université de Recherche Paris Sciences et Lettres, Paris, France
  • 3. Swiss Center for Affective Science, Faculty of Psychology and Educational Sciences, University of Geneva, Geneva, Switzerland
  • 4. Paris-Jourdan Science Economiques CNRS UMR8545, Paris School of Economics, Paris, France

Description

In decision-making, option values can arise from past experiences with rewards and punishments (experiential) or from explicit descriptions of outcomes and probabilities (descriptive). While most studies have examined experiential and descriptive options separately, real-world decisions, such as deciding between a restaurant we visited before and one known through online reviews, often entail choosing between the two. Recently, Garcia et al. (2023) revealed that experiential options are systematically neglected in such hybrid choices. Yet, the cause of this experiential neglect remains unknown. To test whether this phenomenon results from representational differences between experiential and descriptive options, we manipulated the learning architecture, test-phase feedback, stimulus format, and feedback format over five experiments (n = 375). Additionally, participants showed reduced sensitivity to experiential losses relative to gains. Experiential neglect persisted across all conditions, eliminating these hypotheses as explanations of this phenomenon. Rather, we propose that neglect may stem from retrieval-based uncertainty in experiential valuation. Together, these findings provide evidence for a cognitive constraint on the integration of experiential and described values in decision-making.

Files

PiogerGarcia_Experiential_Neglect_Preprint.pdf

Files (6.2 MB)

Name Size
md5:d2a3ef4bb7090ecacff889a79e5d77d2
1.6 MB Preview Download
md5:3c78ead7ff68f37d421d3d90d040e7d0
4.6 MB Preview Download