Stereotypes and Belief Updating.

This article investigates how gender stereotypes influence the updating of self-assessed ability beliefs after receiving noisy, task-specific feedback across domains varying in gender-typing. Using a controlled experiment with 1,989 U.S.-based participants completing tests in eight domains classified as male-typed or female-typed, the study elicits incentivized prior and posterior beliefs about absolute and relative performance. Results show that gender stereotypes significantly predict both prior and posterior beliefs, with men and women updating their beliefs more positively in gender-congruent domains despite receiving highly informative feedback, deviating systematically from Bayesian updating. The study finds that more precise feedback does not reduce reliance on stereotypes, and that responsiveness to good news is greater in gender-congruent domains for both genders, suggesting that stereotypes shape how individuals incorporate new information and contribute to the persistence of gender gaps in self-confidence. These findings have implications for policies aimed at closing gender confidence gaps, indicating that feedback alone may be insufficient without addressing underlying stereotype effects.