Imagine All the People: Investigating People’s Perceptual Biases as They Pertain to Age, Race, and Gender

Abstract

Typically, perceptual biases are studied by investigating how people respond to written scenarios, without considering the mental representations people form while reading these descriptions. This paper provides a novel approach to face perception research by looking at people’s mental representations of strangers and aims to determine whether current ways of classifying people into definite race, age, and gender categories were accurate or needed to be rethought. Specifically, participants digitally reproduced the faces they imagined while reading different scenarios where strangers were described only by race, age, and gender (N = 76). Subsequently, a different set of participants rated these faces on various traits (N = 1024). In the first part of the study, participants created 9 faces from written descriptions of strangers, the last of which included information about criminal history. In the second part, participants rated these faces on dimensions of attractiveness, trustworthiness, intelligence, and physical strength for faces in the non-crime condition, and on dimensions of threat, criminality, and attractiveness for the crime condition. Linear regression models showed that age, race, and gender had various effects on scores on different dimensions, as well as on within-group variance. For instance, older faces were awarded lower attractiveness ratings than younger faces overall, an effect which was also moderated by race, with older age being less predictive of attractiveness ratings for Black faces. Furthermore, there was significantly less variability in attractiveness ratings for Black faces than White faces. Overall, this study revealed that stereotypes do not always adhere to clear-cut categories of race, age, and gender, suggesting that they may be applied somewhat dimensionally rather than categorically.