Cross-Race Facial Recognition

Abstract

Two studies were conducted to increase our knowledge of cross-race recognition of White and Oriental faces and to test the hypothesis that the "cross-race effect" (inferior facial recognition of other races) is due to lack of contact with the other race. In Experiment 1, White (n = 60) and Oriental (n = 60) university students in Canada attempted to recognize White and Oriental faces in a standard facial recognition paradigin. Although the cross-race effect was replicated for false alarms and d', neither perceived similarity nor self-rated contact predicted recognition accuracy. In Experiment 2, White (n = 92) and Oriental (n = 115) students from Singapore and Canada were tested. Contact with Whites and Orientals differed significantly for students in Singapore versus Canada but was not related to facial recognition even though the cross-race effect was replicated. On average, 6 predicted effects of the cross-race effect from the two experiments accounted for 10.83% of the variance, whereas the 18 predicted effects based on the contact hypothesis on average accounted for only 0.89% of the variance in facial recognition. The "contact hypothesis" is not a viable explanation of the results in studies of cross-race facial recognition. The cross-race effect remains unexplained.