Abstract
The other-race effect (ORE: better recognition of own-race faces) is typically studied by presenting faces sequentially to participants and testing recognition using an old/new task. In the real world, people encounter multiple faces simultaneously and in complex scenes and so other-race faces compete for attention. Preferential attention to own-race faces on a daily basis and increased uncertainty when recognizing others in the real world may magnify the ORE in comparison to the relatively small effects typically found in the lab. We compared own- versus other-race face recognition under two different study and test conditions (n = 20 Caucasian participants per group). Participants studied 32 faces (16 Chinese; 16 Caucasian) sequentially (2s per face) or in arrays (shown for 24s) comprising eight faces (four Chinese) and multiple household objects. Recognition was tested using an old/new sequential-presentation task or a "lineup" task in which the proportion of old versus new faces varied across trials. We hypothesized accuracy would be higher in the traditional task but that the ORE would be larger in the array task. The d’ values were higher for own-race faces than other-race faces (p<.001). The magnitude of the ORE varied as a function of learning style (p=.03) but not testing style (p=.72). Surprisingly, accuracy was higher when faces were presented sequentially than in arrays and this effect was larger for own-race faces. Increasing presentation times of the arrays (40s; n = 20) did not alter this pattern of results. Our results suggest that even own-race faces are hard to recognize when they compete for attention with other stimuli (Mean d’ = .67) and that perceptual expertise for own-race faces may be most evident under ideal (i.e., sequential) viewing conditions.
Meeting abstract presented at VSS 2013