Purchase this article with an account.
Ipek Oruc, Xiaoyue M Guo, Jason J S Barton; A contrast-based adaptation study of the contribution of gender to face representations. Journal of Vision 2009;9(8):455. doi: https://doi.org/10.1167/9.8.455.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Face adaptation has been used recently to infer the organization of representation space for faces. Previous face adaptation studies have suggested both distinct and common underlying mechanisms for faces of different gender. We used a new technique that measures the effect of an adapting face on recognition contrast thresholds for subsequent test faces, to determine if adaptation aftereffects are influenced by gender.
We used two female and two male faces in a four-alternative forced-choice paradigm. An ideal observer analysis was used to select a set of four faces in which the physical differences between faces of the same gender were equivalent to the differences between faces of a different gender. During a 100ms adapting period subjects viewed one of five stimuli: one of the four faces, or a blank for the unadapted baseline condition. After a mask, a test face was presented for 150ms, following which the subject indicated which one of the four faces they saw. We used a staircase procedure to determine recognition thresholds for all 20 adapt-test pairs (5 adaptor stimuli x 4 test faces). Nine subjects participated.
The three main conditions were same-face, same-gender (different face of the same gender), and different-gender. We calculated threshold elevation ratios for each main condition by dividing thresholds by the corresponding baseline threshold. We found a significant main effect of condition (Kruskal-Wallis one-way ANOVA, psame-face decreased recognition thresholds, indicating facilitation, while adapting to a different face elevated thresholds. Adapting to a different-gender face increased thresholds more than adapting to a same-gender face, despite our controlling for physical similarity. This result suggests that gender is a factor in representational face-space that cannot be accounted for by simple physical resemblance.
This PDF is available to Subscribers Only