Abstract
Last year at VSS, we showed that the same spatial frequencies (SFs) are used for the identification of upright and inverted inner facial features (Abstract #153). Here, we report three follow-up experiments based on the same SF Bubbles technique to shed light on the relationship between the face inversion effect (Yin, 1969) and the composite face effect (Young, Hellawell, & Hay, 1987). In Experiment 1, we replicated our previous findings on the face inversion effect in a 10-choice identification task with 300 trials per orientation and per observer and with 20 faces from the set of Goffaux and Rossion (2006) revealed through an elliptical aperture hiding contour information—the same SFs were used to identify upright and inverted faces. In Experiment 2, we displayed the faces of Experiment 1 with contour information. For upright face identification, we replicated our previous results for inverted faces, however, the use of SFs was shifted toward lower SFs. Intriguingly, this shift is in the opposite direction to that predicted by Goffaux and Rossion (2006) who found that holistic processing is largely supported by low SFs. In Experiment 3, we re-examined SF tuning in the composite face paradigm of Goffaux and Rossion (2006) using the SF Bubbles technique. Preliminary results confirm and extend their results. In sum, holistic processing—as indexed by the composite face effect—and face identification appear to be mediated by different SFs.