Rapid Fear Detection Predicts Prosocial Giving

Abstract Prosociality is fundamental to human social life, yet individuals vary widely in their willingness to incur costs for others. What drives this variation? We tested four preregistered hypotheses about whether the speed of detecting specific social stimuli (faces vs. cars, infant vs. elderly faces, fearful vs. angry faces, trustworthy vs. untrustworthy faces) is associated with prosocial giving. Participants (N = 200) completed visual detection tasks, personality assessments, a Director Task measuring Theory of Mind, and two economic games: the Dictator Game, indexing costly giving in the absence of strategic pressure, and the Ultimatum Game, indexing giving under strategic constraints. One hypothesis received partial support: faster detection of fearful faces was associated with greater giving, an effect that was most robust in the Dictator Game and was significant in the primary analyses. Robustness checks confirmed the Dictator Game finding across analytic approaches, while the Ultimatum Game effect was attenuated in analyses using robust standard errors and ordinal logistic regression. Neither detection of other social stimuli, Big Five personality traits, facial width-to-height ratio, nor Theory of Mind performance reliably predicted giving. These findings suggest that absolute speed of detecting distress signals reliably predicted prosocial giving, that this effect did not reflect general processing speed, and that general social attunement, vulnerability sensitivity, and higher-order social cognition did not reliably predict giving.