Abstract
Matching the identity of unfamiliar faces is highly error-prone, especially when the face is a different race to the viewer. This is referred to as the other-race effect (ORE). Recent research has tested methods for improving unfamiliar face matching performance and found that exposure to within-person variability (the way a face varies across multiple encounters) improves the identification of unfamiliar own-race faces. The current study investigates whether exposure to within-person variability can improve the identification of other-race faces as well as reduce the ORE. Using a cross-over design (N= 105), Caucasian-Australian and Asian-Singaporean participants completed a simultaneous matching task with Caucasian and Asian faces. Participants viewed a target image with either one other ambient image or an array of three ambient images and judged whether the identity in the array was the same or different to the target image identity. Results showed that viewing multiple images improved matching accuracy for own- and other-race faces in trials where the target image identity was the same as the image array. However, multiple images decreased accuracy in trials where the target image was a different identity. Notably, although multiple images improved accuracy for matching other-race faces, the improvement was not different from that for own-race faces. That is, the ORE was not reduced. This pattern was also shown in the analysis of signal detection measures. These findings highlight the robust nature of the ORE and provide insight into the formation of familiar face representations.