Perspective taking reduces intergroup bias in visual representations of faces
Related Experiment Video
Updated: Nov 1, 2025

Post-Movie Subliminal Measurement PMSM, for Investigating Implicit Social Bias
Published on: February 29, 2020
Ryan J Hutchings1, Austin J Simpson1, Jeffrey W Sherman1
1University of California, Davis, CA, USA.
Abstract:
Intergroup biases shape most aspects of person construal, including lower-level visual representations of group members' faces. Specifically, ingroup members' faces tend to be represented more positively than outgroup members' faces. Here, we used a reverse-correlation paradigm to test whether engaging in perspective taking (i.e., actively imagining another person's mental states) can reduce these biased visual representations. In an initial image-generation experiment, participants were randomly assigned to a minimal group and then composed a narrative essay about an ingroup or an outgroup target person, either while adopting the person's perspective or while following control instructions. Afterward, they generated an image of the person's face in a reverse-correlation image-classification task. Subsequent image-assessment experiments using an explicit rating task, a sequential priming task, and an economic trust game with separate samples of participants revealed that ingroup faces elicited more likability and trustworthiness than did outgroup faces. Importantly, this pattern of intergroup bias was consistently weaker in faces created by perspective takers. Additional image-assessment experiments identified the mouth (i.e., smiling cues) as a critical facial region wherein the interactive effects of group membership and perspective taking emerged. These findings provide initial evidence that perspective taking may be an effective strategy for attenuating, though not for eliminating, intergroup biases in visual representations of what group members look like.
More Related Videos
Related Concept Videos
Gestalt Principles of Perception
Halo Effect
Actor-Observer Effect
Stereotype Content Model
Depth Perception and Spatial Vision
Stereotypes, Prejudice, and Discrimination

