In face perception, reducing visual input greatly increases perceived attractiveness; left/right half faces look far more attractive than bilaterally symmetric whole faces

Face perception loves a challenge: Less information sparks more attraction. Javid Sadr, Lauren Krowicki. Vision Research, Volume 157, April 2019, Pages 61-83. https://doi.org/10.1016/j.visres.2019.01.009

Highlights
•    In face perception, reducing visual input greatly increases perceived attractiveness.
•    This “partial information effect” occurs with blur, contrast reduction, and occlusion.
•    Left/right half faces look far more attractive than bilaterally symmetric whole faces.
•    There are no male/female differences in this “less is more” enhancement effect.

Abstract: Examining hedonic questions of processing fluency, objective stimulus clarity, and goodness-of-fit in face perception, across three experiments (blur, contrast, occlusion) in which subjects performed the simple, natural task of rank-ordering faces by attractiveness, we find a very consistent and powerful effect of reduced visual input increasing perceived attractiveness. As images of faces are blurred (i.e., as higher spatial frequencies are lost, mimicking at-a-distance, eccentric, or otherwise unaccommodated viewing, tested down to roughly 6 cycles across the face), reduced in contrast (linearly, down to 33% of the original image’s), and even half-occluded, the viewer’s impression of the faces’ attractiveness, relative to non- or less-degraded faces, is greatly enhanced. In this regard, the blur manipulation exhibits a classic exponential profile, the contrast manipulation follows a simple linear trend. Given the far superior attractiveness of half-occluded faces, which have no symmetry whatsoever, we also see that it may be incorrect to claim that facial symmetry is attractive and perhaps more accurate that asymmetry may be unattractive. As tested with a total of 200 novel female faces over three experiments, we find absolutely no male/female differences in this “partial information effect” of enhanced subjective attraction, nor do we find differences across the repetition of the task through to a second block of trials in which the faces are re-encountered and no longer novel. Finally, whereas objective stimulus quality is reduced, we suggest a positive hedonic experience arises as a subjective phenomenological index of enhanced perceptual goodness-of-fit, counter-intuitively facilitated by what may be stimulus-distilling image-level manipulations.