An Adult Developmental Approach to Perceived Facial Attractiveness and Distinctiveness

Abstract

Attractiveness and distinctiveness constitute facial features with high biological and social relevance. Bringing a developmental perspective to research on social-cognitive face perception, we used a large set of faces taken from the FACES Lifespan Database to examine effects of face and perceiver characteristics on subjective evaluations of attractiveness and distinctiveness in young (20-31 years), middle-aged (44-55 years), and older (70-81 years) men and women. We report novel findings supporting variations by face and perceiver age, in interaction with gender and emotion: although older and middle-aged compared to young perceivers generally rated faces of all ages as more attractive, young perceivers gave relatively higher attractiveness ratings to young compared to middle-aged and older faces. Controlling for variations in attractiveness, older compared to young faces were viewed as more distinctive by young and middle-aged perceivers. Age affected attractiveness more negatively for female than male faces. Furthermore, happy faces were rated as most attractive, while disgusted faces were rated as least attractive, particularly so by middle-aged and older perceivers and for young and female faces. Perceivers largely agreed on distinctiveness ratings for neutral and happy emotions, but older and middle-aged compared to young perceivers rated faces displaying negative emotions as more distinctive. These findings underscore the importance of a lifespan perspective on perception of facial characteristics and suggest possible effects of age on goal-directed perception, social motivation, and in-group bias. This publication makes available picture-specific normative data for experimental stimulus selection.

Keywords: age; attractiveness; cross-classified random effects analysis; distinctiveness; emotion; faces.

FIGURE 1

Illustration of the rating procedure. During each session, participants rated the attractiveness and the distinctiveness of different face pictures by adjusting a slider on a scale ranging from 0 to 100. Written informed consent has been obtained from depicted individuals for the publication of the images.

FIGURE 2

Predicted marginal means for attractiveness and distinctiveness for specified two-way interactions pertaining to age and gender: (A) Age of Perceiver × Age of Face, for attractiveness. (B) Age of Perceiver × Age of Face, for distinctiveness. (C) Age of Perceiver × Gender of Perceiver, for attractiveness. (D) Age of Perceiver × Gender of Perceiver, for distinctiveness. (E) Age of Face × Gender of Face, for attractiveness. (F) Age of Face × Gender of Face, for distinctiveness. Error bars indicate 95% confidence intervals for the point predictions, corrected for multiple comparisons (Bonferroni). P = Perceiver; F = Faces.

FIGURE 3

Predicted marginal means for attractiveness and distinctiveness for specified two-way interactions pertaining to age, gender, and emotion of face: (A) Age of Perceiver × Emotion of Face, for attractiveness. (B) Age of Perceiver × Emotion of Face, for distinctiveness. (C) Age of Face × Emotion of Face, for attractiveness. (D) Age of Face × Emotion of Face, for distinctiveness. Error bars indicate 95% confidence intervals for the point predictions, corrected for multiple comparisons (Bonferroni). P = Perceiver; F = Faces.

FIGURE 4

Variations in the relation between attractiveness and distinctiveness ratings by age of perceiver and face: (A) Young Participants, Young Faces. (B) Young Participants, Middle-Aged Faces. (C) Young Participants, Older Faces. (D) Middle-Aged Participants, Young Faces. (E) Middle-Aged Participants, Middle-Aged Faces. (F) Middle-Aged Participants, Older Faces. (G) Older Participants, Young Faces. (H) Older Participants, Middle-Aged Faces. (I) Older Participants, Older Faces. Black lines show expected values of attractiveness, conditional on distinctiveness. Shaded areas indicate 95% confidence regions. Results were obtained from separate fits of multilevel restricted cubic spline models (Eq. 1). P = Perceiver; F = Faces.

FIGURE 5

Variations in the relation between attractiveness and distinctiveness ratings by emotion of face: (A) Neutral. (B) Happy. (C) Sad. (D) Angry. (E) Fearful. (F) Disgusted. Black lines show expected values of attractiveness, conditional on distinctiveness. Shaded areas indicate 95% confidence regions. Results were obtained from separate fits of multilevel restricted cubic spline models (Eq. 1). P = Perceiver; F = Faces.