Abstract
Recent evidence suggests that stronger holistic processing of own-race faces may underlie the well-established own-race advantage for face memory. Across a number of studies Caucasian participants have demonstrated significantly larger holistic processing effects for Caucasian over Asian faces. However, Asian participants have shown a consistently different pattern: similar sized effects for both Asian and Caucasian faces. Here, Experiment 1 replicated these previous findings using Tanaka & Farah's (1993) part-whole task. Caucasian Australians displayed a significantly larger whole face advantage for Caucasian than Asian faces, while Hong Kong Chinese showed no race-of-face differences in holistic processing. Additionally, we included an inverted face condition to investigate the possibility that results for Asian participants reflect a domain-general global processing bias. Results suggest that Asian participants, unlike Caucasian participants, use face-specific holistic mechanisms to process both own- and other-race upright faces. Experiment 2 used the part-whole task in conjunction with Palermo & Rhodes' (2002) flanker task. In the study phase of each trial the part-whole target was flanked by two faces. Participants performed an across-viewpoint identity-matching task (same/different) on the flanker faces while simultaneously encoding the central target. The presence of holistic processing for the flanker task is argued to disrupt holistic processing of the targets. For Caucasian participants holistic processing for own-race targets was reduced only when the flanker faces were Caucasian. For Asian participants both Asian and Caucasian flankers interfered with holistic processing for own-race targets, however no reduction in holistic processing for other-race targets was observed for either race of flankers. Results for this task again support a difference between Asian and Caucasian participants in holistic processing for other-race face. Together these results suggest that, at least for Asian participants, differences in the strength of holistic processing do not explain differences in memory for own- and other-race faces.
This study was supported by a grant from the Hong Kong Research Grants Council (HKU744209) to William Hayward.