Culture and trait inferences from facial cues

Prior research showed that people make inferences about personality traits based on facial features and that there is cross-cultural consensus concerning such face-based trait inferences. The current research tested whether there are cultural differences in the extent to which people ascribe traits to individuals based on facial features. We built Caucasian and Asian faces using a data-driven statistical model that represents two fundamental social dimensions—trustworthiness and dominance traits—and generated faces that vary on each dimension, ranging from − 4.5 SD to + 4.5 SD. By asking European American and Korean participants to judge these faces, we demonstrated that although trait evaluations tracked the trait intensity predicted by the computer models in both cultures, European American participants inferred more extreme traits from faces than did Korean participants. Cultural differences in the extent of trait ascription were partly explained by attention to situational information. In addition, participants made more differentiated trait ascription based on their own-race faces than other-race faces; this was more pronounced among European American participants.