Research Highlights

A deeper look at facial perception

Published online 11 January 2021

Study questions the universality of existing theories of face evaluation by collecting data from multiple world regions.

Rieko Kawabata

Getty Images
How we evaluate faces impacts many aspects of human behaviour. We are more likely to interact socioeconomically with people whose faces we perceive as trustworthy, and are drawn to those we view as attractive. 

Over the past decade, behavioural scientists have used Oosterhof and Todorov’s valence–dominance model as a framework for face evaluation. The model takes into account 13 traits, including attractiveness, intelligence, confidence, aggressiveness, meanness, emotional stability and even weirdness. Based on ratings of these traits, the model proposes that two factors underpin how we form impressions of others: valence (how trustworthy the individual appears) and dominance (how physically strong they seem to be). So far, the model has largely been developed and tested in Western regions. 

Now, a study conducted by more than 100 psychology researchers around the world, including authors from Morocco and the United Arab Emirates, has sought to address whether the valence–dominance model may similarly apply to other world regions. 

The study involved collecting data from 11 regions (Africa, Asia, Australia and New Zealand, Central America and Mexico, Eastern Europe, the Middle East, the United States and Canada, Scandinavia, South America, the United Kingdom and Western Europe), 41 countries and more than 11,000 participants. In the Middle East, data was collected from Iran, Israel and Turkey.

Participants were asked to rate the 13 different traits using a scale ranging from one to nine for 120 faces photographed with a neutral expression representing 30 black (15 male, 15 female), 30 white (15 male, 15 female), 30 Asian (15 male, 15 female) and 30 Latin (15 male, 15 female) individuals. Each participant rated the 120 faces twice and the results were averaged to ensure reliability of the ratings.

The study had two key findings. When using Oosterhof and Todorov’s original method of analysis, the valence–dominance model was found to be broadly applicable across all world regions. In contrast, regional differences were revealed when using an alternative analysis that factored in common criticisms of the original method.

This suggests there is greater complexity in the way that trait ratings correlate among different cultures than can be explained by the original model. By broadening the focus of research beyond Western regions, the researchers say the study provides a basis for expanding or refining existing theories. More work would be needed to understand differences in the impression formation process at the regional level.


Jones, B. C. et al. To which world regions does the valence–dominance model of social perception apply? Nat. Hum. Behav. (2020).