It's All Relative: Associations Of Facial Proportionality, Attractiveness, And Character Traits In 597 Individuals
Dillan F. Villavisanis, BA1, Clifford I. Workman, PhD2, Daniel Y. Cho, MD PhD1, Zachary D. Zapatero, BS1, Connor S. Wagner, BS2, Liana Cheung, MBBS1, Jessica D. Blum, MSc1, Scott P. Bartlett, MD1, Jordan W. Swanson, MD MSc1, Anjan Chatterjee, MD2, Jesse A. Taylor, MD1.
1Children's Hospital of Philadelphia, Philadelphia, PA, USA, 2Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA.
PURPOSE: Facial proportionality has known associations with perceptions of sociability, intelligence, and health; however, many previous studies used small sample sizes or artificial facial renderings. Using a large dataset, this study aimed to 1) determine the association of proportionality with attractiveness and character traits 2) determine differences in attractiveness and character ratings between "anomalous" and "typical" faces using a large dataset.
METHODS: 597 individuals were included from Chicago Face Database, a domain of facial measurements and metrics by 1,087 raters. The equation "Proportionality" (Figure1A) quantified horizontal proportionality: "0" indicated perfect proportionality and more negative scores indicated disproportionality. Individuals were categorized as "anomalous" (ie: jaw asymmetry, scars) or "typical" by two reviewers.
RESULTS: Spearman's correlations revealed proportionality was associated with attractiveness (ρ=0.292, p<0.001) and trustworthiness (ρ=0.193, p<0.001), and disproportionality with anger (ρ=0.132, p=0.001), dominance (ρ=0.259, p<0.001), and threateningness (ρ=0.234, p<0.001). Mann-Whitney U tests revealed the "typical" cohort had higher levels/ratings of proportionality (-13.89 vs. -15.26, p=0.008), attractiveness (3.43 vs. 2.95, p<0.001), and trustworthiness (3.49 vs. 3.34, p<0.001). Multivariate logistic regression models demonstrated interactions between proportionality and attractiveness predicted presence of facial anomalies (Figure1B-C; OR=1.056, 95% CI=0.008-0.102; β=0.055, SE=0.024, z=2.285, p=0.022).
CONCLUSION: This study demonstrates associations of facial proportionality with ratings of attractiveness and character traits. Additionally, proportionality and attractiveness both play a role in predicting the presence of facial anomalies.
Back to 2022 Posters