Anjana Lakshmi

← Back to Research

The Hyper-Typicality Effect: the Typical Face is Not Average

Under review

Collaborators: Bernd Wittenbrink, Josh Correll, Debbie Ma

Abstract

Perceived group typicality of faces plays a central role in social categorization and stereotyping. It is commonly assumed that the most typical group members are those whose facial features approximate the group average. The present research tests this assumption by examining how perceived ethnic typicality varies as faces deviate from the group average. Across three studies (N = 1,340), Black and White male faces were rated for how typical they appeared of their respective social groups as a function of their position along a group-differentiating dimension of facial variation. Studies 1a and 1b used a computational approach to capture naturally occurring facial variation in high-dimensional face spaces. Results showed that faces with increasingly exaggerated group-diagnostic features were judged as more group typical than average faces, particularly for Black faces. Study 2 experimentally manipulated faces to vary systematically their group-differentiating features. For both Black and White faces, perceived group typicality again peaked for faces with more extreme group-differentiating features, although excessive extremity ultimately reduced typicality judgments. Participant ethnicity moderated these effects, with greater accentuation of group-differentiating features observed in judgments of outgroup faces. Together, these findings demonstrate that perceived group typicality of faces is not anchored at the average but is systematically displaced toward exaggerated exemplars. The results challenge average-based models of typicality and advance a more precise understanding of how facial features shape social category representations.