Do infants attribute false beliefs to a visible agent with no visual access?

Diane Poulin-Dubois, Elizabeth J. Goldman

Infants’ ability to attribute false beliefs has been widely studied using the violation-of-expectation (VOE) paradigm. However, the robustness of these findings remains debated, particularly when methodological variations are introduced. This study examined whether 16-month-old infants attribute false beliefs to agents who are physically present but lack perceptual access to a change in object location. In two conceptual replications of the false belief VOE paradigm, we manipulated the agent’s visual access during the belief-induction trial: the agent wore an opaque veil (Experiment 1), or the agent turned her back (Experiment 2). Results showed no significant differences in looking time between the Congruent and Incongruent conditions, and Bayesian analyses provided moderate support for the null hypothesis. These findings suggest that infants may not flexibly attribute false beliefs across contexts, raising questions about the generalizability and robustness of early false belief understanding.