Abstract
How do people judge groups in terms of attractiveness? We investigated whether people compute and utilize the average attractiveness of group members. In Experiment 1, two groups of participants rated the attractiveness of 240 face images. One group rated one by one and the other group rated set by set (consists of 4 faces). We found that the attractiveness rating for a set of four faces was equivalent to the arithmetic mean of ratings for each face in a set. In Experiment 2, we modulated the attractiveness variance within a set and asked participants to rate the attractiveness of a set as a whole. The attractiveness rating of a set was equivalent to the arithmetic mean of ratings of its members, only when the attractiveness variance was low. These patterns were maintained whether a set of faces was presented for either 500ms or 2000ms. These results suggest that the visual system efficiently computes the average attractiveness of group members, when the faces are similarly attractive. Finally, Experiment 3 investigated whether participants' judgments on the average attractiveness of faces in a group were biased towards the most attractive face in the group. Here, participants were asked to detect a probe presented after each group of faces. Reaction time (RT) to the probe replacing the location of the most attractive face did not significantly differ from the RT to the probe replacing the locations of the other faces, suggesting that the visual system averages all faces in a group with equal weights.
Meeting abstract presented at VSS 2016