Perceived femininity and masculinity contribute independently to facial impressions

Hester, Neil and Jones, Benedict C. and Hehman, Eric (2021) Perceived femininity and masculinity contribute independently to facial impressions. Journal of Experimental Psychology: General, 150 (6). 1147–1164. ISSN 0096-3445 (

[thumbnail of Hester-etal-JEPG-2020-Perceived-femininity-and-masculinity-contribute-independently-to-facial-impressions]
Text. Filename: Hester_etal_JEPG_2020_Perceived_femininity_and_masculinity_contribute_independently_to_facial_impressions.pdf
Accepted Author Manuscript

Download (2MB)| Preview


In person perception research, femininity and masculinity are regularly conceived as 2 ends of 1 bipolar dimension. This unidimensional understanding permeates work on facial impressions, gender diagnosticity, and perceptions of LGBTQ individuals, but it is perhaps most prominent in evolutionary work suggesting that sexually dimorphic facial features (which vary along a female–male continuum) correspond directly with subjective ratings of femininity and masculinity, which in turn predict ratings of traits such as attractiveness. In this paper, we analyze 2 large face databases (the Chicago and Bogazici Face Databases) to demonstrate that femininity and masculinity are distinct dimensions in person perception. We also evaluate key theoretical assumptions surrounding femininity and masculinity in evolutionary theories of face perception. We find that sexually dimorphic features weakly correlate with each other and typically explain just 10–20% of variance in subjective ratings of femininity and masculinity. Femininity and masculinity each explain unique variance in trait ratings of attractiveness, dominance, trustworthiness, and threat. Femininity and masculinity also interact to explain unique variance in these traits, revealing facial androgyny as a novel phenomenon. We propose a new theoretical model explaining the link between biology, facial features, perceived femininity and masculinity, and trait ratings. Our findings broadly suggest that concepts that are "opposites" semantically cannot necessarily be assumed to be psychological opposites.