THE EFFECTS OF RACE ON FACE RECOGNITION

Three experiments explored the effects of race on face recognition. In Experiment 1, white and black subjects saw slides of own- and other-race faces in upright and inverted orientations. All subjects showed an "own-race effect:" superior performance on own-race faces, for upright faces. White, but not black, subjects showed the same effect for inverted faces. All subjects performed poorly on inverted black faces. False alarm rates were highest where sensitivity was lowest: for other-race and inverted faces. The results fail to support the notion that lack of experience with other-race faces is responsible for the own-race effect. In Experiment 2, white and black subjects judged whether two simultaneously-presented slides of the same face were in the same left-right orientation. Performance was higher on white faces, but there was no own-race effect. Subjects made more Same than Different judgments, especially for black faces. The results failed to support the notion that the own-race effect is due to impoverished perception of other-race faces and suggested that the poor performance on inverted black faces might be due to their greater symmetry. Experiment 3 differed from Experiment 2 only in that the judgment was made from memory of a standard. Performance was poor, but among those scoring above chance, all of the findings of Experiment 2 were replicated, showing that the difference between black and white faces persists in a memory task. However, there was no correlation between percent correct (for a given face) in the left-right orientation judgment and inverted recognition, and therefore no evidence of a connection between the poor performance on black faces in Experiments 1 and 2. Race emerged as a variable that effects recognition performance reliably but somewhat weakly. It was concluded that motivation is probably involved in face recognition, and that the own-race effect reflects memory rather than perception of faces.