Eye gaze is not inversion-proof: A robust, sex-invariant gaze inversion effect

Humans are adept at distinguishing individual faces, yet inversion dramatically impairs this ability. This face inversion effect is remarkably robust across observers, but evidence is mixed as to whether inversion also impairs the perception of facial parts, particularly the eye region. Some studies have shown that featural processing is preserved or even enhanced when faces are inverted, whereas others have reported clear inversion-related impairments in feature-based judgements. These mixed findings may reflect limited statistical power, unbalanced participant sex ratios, and heterogeneous task designs. To address these issues, we examined how strongly face inversion affects sensitivity to gaze direction in a well-powered, sex-balanced sample. A total of 190 participants judged whether the eyes in briefly presented upright or inverted faces were looking directly at them or not. Inversion reliably reduced sensitivity to gaze direction, yielding a medium-to-large effect size. Females showed modestly higher overall sensitivity than males (a small-to-medium effect), whereas the inversion effect was highly similar for females and males. These findings show that brief gaze judgements are not immune to inversion, even in a task that could in principle be based largely on eye-region information. They provide quantitative constraints that models of gaze perception should accommodate, including the largely sex-invariant inversion effect.