Eye movements for scrambled faces

Abstract

We now have considerable evidence on the nature of face perception, and the way in which observers acquire information from intact faces in order to make judgments (e.g., identity, sex, age, ethnicity) about them. However, recent work has shown that faces can often be successfully recognized on the basis of individual features rather than the whole, intact face. Therefore, the goal of the current study was to investigate eye movements when participants were viewing sets of facial features that had been scrambled from their original configuration. Participants viewed scrambled and intact faces, in the context of a recognition memory test (where half the test items had been studied and half were new). During both study and test trials, fixation position was monitored. When presented with scrambled stimuli, fixations were largely centered on the two eyes, with relatively few fixations to the mouth, nose, or other features. With intact stimuli, however, fixations showed a different pattern, with more fixations to the nose and mouth (as well as to the eyes). We attribute this difference in eye movement patterns between intact and scrambled faces to the influence of the overall facial configuration in the former case. When facial features appear in the context of the intact facial configuration, the visual system is able to efficiently acquire information from across the whole face. Once the features are scrambled, however, observers appear to use a much more restricted focus of attention, which they position mainly on the eyes. These results also suggest that observers use a relatively small number of features as the basis for recognition decisions about scrambled faces.