Fooling Yourself: how narratives shape beliefs

2026-07-06

AI Summary

This study investigates how non-diagnostic narrative cues interfere with evidence-based Bayesian belief updating in decision-making. In a controlled laboratory experiment, participants were randomly assigned to narrative, abstract, or no-information control conditions. Combining reduced-form estimation with subject-level classification, the research quantifies the impact of irrelevant narrative details on belief dynamics. The findings reveal that non-diagnostic narrative cues systematically bias beliefs toward maximal uncertainty, significantly slowing belief convergence and increasing the amount of information required to reach a decision. This effect attenuates in abstract tasks and disappears entirely under no-information conditions. The results uncover a subtle yet systematic mechanism by which narrative structure disrupts rational inference and provide a quantitative measure of the resulting loss in decision efficiency.

Abstract

Decision-makers usually receive information through narratives that combine diagnostic evidence with nondiagnostic details. In a laboratory experiment, we study how such nondiagnostic clues affect belief updating. Participants repeatedly report beliefs in a Bayesian inference task within a narrative context. Reduced-form estimates and subject-level classifications show that nondiagnostic narrative clues systematically induce belief revision toward maximal uncertainty, weakening previously accumulated diagnostic evidence. This effect is weaker in an equivalent abstract context and disappears in an identical narrative task when nondiagnostic clues are replaced by no-information messages. We assess the economic significance of such return to uncertainty by showing that it delays belief convergence, thereby increasing the amount of information agents require before making a decision.

Problem

• narratives
• belief updating
• nondiagnostic clues
• Bayesian inference
• uncertainty

Innovation

• narrative bias
• belief updating
• nondiagnostic information
• Bayesian inference
• uncertainty