Abstract
Caucasians are biased to perceive angry expressions in Black faces, resulting both in greater accuracy for recognition of angry Black faces (Albright et al., 2006) and a tendency to misinterpret Black neutral expressions as angry (Maner et al., 2005). Here, we investigated how rapidly other-race effects for emotional expression emerge by manipulating exposure duration of emotional Black and Caucasian faces masked by happy faces before and after target presentation. Target emotions were neutral, angry, disgust, fear and sad. Target exposure durations were 50, 100, 150, 200 ms. 17 Black and 28 Caucasian participants were tested. Chi Square analysis revealed that in general, Caucasians were more likely to respond "angry" to Black than Caucasian faces (p = .014), whereas Caucasians were more likely to respond "fearful" to Caucasian than Black faces (p < .001). Black participants were more likely to respond "sad" to Caucasian than Black faces (p = .05). Analysis of the angry expression condition revealed greater accuracy for Black than Caucasian faces only for Caucasian participants. This effect emerged in the 100 ms exposure duration, and increased linearly as exposure duration increased, F(1, 27) = 7.329, η2 = .213, p = .012. Interestingly, both Black and Caucasian participants were equally more likely to respond "angry" to Black neutral faces than Caucasian neutral faces, and this bias also emerged at 100 ms and increased linearly with exposure duration. However, by 200 ms, the bias disappeared for Black participants (bias = 0) whereas it continued to increase for Caucasian participants (36% more likely to respond Angry to Black neutral faces than Caucasian neutral faces). These results suggest that both Black and Caucasian participants have an angry-Black face bias, but that the processes underlying the biases in the two groups diverge at around 200 ms of exposure duration.
Meeting abstract presented at VSS 2016