Abstract
The history of the study of face perception has been characterized by a search for the behavioral and neurophsyiological signatures of the holistic (configural, gestalt, etc.) processes that are assumed as a hallmark of face perception. Theoretical analyses of these regularities, based on the application of multidimensional signal detection theory (general recognition theory, GRT), have suggested that these regularities can be obtained by a variety of combinations of perceptual and decisional factors. In order to try to disentangle the roles of perceptual and decisional factors contributing to one of these regularities–the composite face effect–Kuefner and Rossion (VSS, 2009) used electrophysiological results (the N170 and the lateralized readiness potential, LRP) to suggest that the composite face effect–and, by extension, the assumed holistic processing of faces–is driven solely by perceptual factors. The goal of the present study was to show that the results of Kuefner and Rossion are actually one of a set of outcomes predicted by GRT. We do this by examining performance of individual observers, all of whom completed three experimental tasks involving the same stimuli. The first was a replication of Kuefner and Rossion's composite face task, allowing examination of critical neurophysiological regularities suggestive of perceptual sources. The second was an implementation of the Eriksen flanker task, allowing explicit consideration of aspects of decisional sources. The third was a complete identification task, in which both perceptual and decisional sources could be assessed. Results illustrate how the set of possibilities for the perceptual and decisional sources of the composite face effect predicted can be expressed and measured in the performance of individual observers.