A data-driven test for cross-cultural differences in face preferences

Zhang, Lingshan and Holzleitner, Iris J. and Lee, Anthony J. and Wang, Hongyi and Han, Chengyang and Fasolt, Vanessa and DeBruine, Lisa M. and Jones, Benedict C. (2019) A data-driven test for cross-cultural differences in face preferences. Perception, 48 (6). pp. 487-499. ISSN 0301-0066 (https://doi.org/10.1177/0301006619849382)

[thumbnail of Zhang-etal-Perception-2019-A-data-driven-test-for-cross-cultural-differences-in-face-preferences]
Preview
 Text. Filename: Zhang_etal_Perception_2019_A_data_driven_test_for_cross_cultural_differences_in_face_preferences.pdf
Accepted Author Manuscript
Download (539kB)| Preview

Abstract

Previous research has shown strong cross-cultural agreement in facial attractiveness judgments. However, these studies all used a theory-driven approach in which responses to specific facial characteristics are compared between cultures. This approach is constrained by the predictions that can be derived from existing theories and can therefore bias impressions of the extent of cross-cultural agreement in face preferences. We directly addressed this problem by using a data-driven, rather than theory-driven, approach to compare facial attractiveness judgments made by Chinese-born participants who were resident in China, Chinese-born participants currently resident in the UK, and UK-born and UK-resident White participants. Analyses of the principal components along which faces naturally varied suggested that Chinese and White UK participants used face information in different ways, at least when judging women’s facial attractiveness. In other words, the data-driven approach used in this study revealed some cross-cultural differences in face preferences that were not apparent in studies using theory-driven approaches.

ORCID iDs

Zhang, Lingshan, Holzleitner, Iris J., Lee, Anthony J., Wang, Hongyi, Han, Chengyang, Fasolt, Vanessa, DeBruine, Lisa M. and Jones, Benedict C. ORCID logoORCID: https://orcid.org/0000-0001-7777-0220;
CORE (COnnecting REpositories)