The effect of race, inversion and encoding activity upon face recognition

Abstract

Recognition of faces has been shown to be more impaired by inversion than recognition of other objects normally only seen upright (Yin 1969).> Experiment 1 explores the possibility that this result is explicable in terms of the familiarity of the recognition tasks rather than a ‘face-specific’ factor. However, a less familiar task (recognizing other race faces) was more disrupted by inversion than recognizing own race faces. In experiment 2, Yin's (1969) finding was replicated using a different view of items at test, a task which is more representative of everyday face recognition. Yin (1970) suggested that the disproportionate effect of inversion may be due to difficulty in perceiving facial expression in an'inverted face. However, in experiment 3, subjects encouraged to make personality judgements on initial viewing of the faces were no more impaired by inversion at test than subjects encouraged to name distinctive physical features. These results imply that the disproportionate effect of inversion upon face recognition cannot be explained in terms of the extra familiarity of the task or the use of identical photographs at test. Furthermore, it appears that the role of facial expression is not sufficient to account for this effect. However, the results of experiment 3 are also discussed in terms of the effectiveness of specific encoding instructions to enhance expression analysis selectively.