Facial Attractiveness

Face stimuli manufactured by prototyping can be used to investigate questions related to facial attractiveness. For example, the experiment on this page tests and refutes the hypothesis that the average female face of a population is perceived as being the most attractive, a theory put forward by Langlois & Roggman (1990).

The experiment is split up into two parts: the first identifying whether highly attractive faces differ in shape from a computed average, and the second making sure that any shape differences are causally related to the perceived attractiveness of the face.

Part 1: Is there a difference between the average and highly attractive face shapes?

Sixty female faces were rated for attractiveness by male and female raters. An `Average' prototype was made by blending 60 faces into the average shape for the population. A prototype was formed with a `High' shape by blending all 60 faces into the average shape of the 15 most attractive faces. The High shape differs from the Average shape (see diagram below).

Fig. 1
The High shape (from the 25% of female faces judged most attractive) is illustrated in red. The Average shape of the population of female faces is also illustrated in blue. (The above image is linked to a larger version.)

Part 2: Is the difference between the average and highly attractive face shapes related to attractiveness?

90% of male and female Caucasian subjects preferred the prototype in the High shape (image B) to the prototype in the Average shape (image A) indicating that the shape differences were causally related to perceived attractiveness. Moreover, caricaturing the difference between High and Average shapes produced an image (C) which 70% of subjects preferred to image B. Thus, the most attractive face shape is not average. These findings generalised across cultures and gender of face.

(A) (B) (C)

(A) Average shape, a prototype made from 60 female faces aged 20-30 without makeup. (B) High shape, the prototype reformed into the average shape of a subset of faces rated highly for attractiveness. (C) Enhanced shape difference, prototype reformed by enhancing the shape differences between images (A) and (B) by 50%.

For further information please see:

Perrett D.I., May K., & Yoshikawa S. (1994) Attractive characteristics of female faces: preference for non-average shape. Nature, Lond. 368, 239-242. abstract

Perrett D.I., Less K.J., Penton Voak I., Rowland D., Yoshikawa S., Burt D.M., Henzil S.P., Castles D.L., & Akamatsu S. (1998) Effects of sexual dimorphism on facial attractiveness. Nature, 394, 884-887. abstract

Images and text Copyright 1995, The Perception Lab, University of St Andrews
Attractiveness / dmb@st-and.ac.uk

Perception lab. School of Psychology




Back to top of page.