| The Guns and Tools task (Payne, 2001) was modified in the current research so that the Black and White face primes displayed emotional expressions. Thus, the design employed was 2 (Prime Ethnicity: Black vs. White) x 3 (Prime Facial Affect: Angry vs. Happy vs. No Emotion) x 2 (Target). Ethnicity and facial affect were crossed and the influences of each on (1) the accurate perception of threat, (2) the misperception of threat, and (3) the automatic vs. controlled processing identified using the process dissociation procedure (PDP; Jacoby, 1991) were examined. Predictions were that both racial cues and affective cues would cause differential patterns of slowed processing, errors, and automatic estimates such that, the negativity associated with skin color and the negativity associated with facial affect would impact the accurate perception of threat. N = 63 Washington University undergraduates participated in this study in exchange for course credit. Regarding slowed processing, there was a 3-way interaction among Ethnicity, Affect and Target, F (2, 92) = 3.74, p < .05, eta 2 = .075. Regarding errors, there were two 2-way interactions between Affect and Target, F (2, 92) = 3.94, p < .05, eta2 = .079, and Ethnicity and Target, (1, 46) = 28.30, p < .001, eta2 = .381. Regarding Automatic estimates, there was a significant interaction between Ethnicity and Affect, F (2, 92) = 15.45, p < .001, eta2 = .251. The data supported predictions; the degree to which responses were influenced by the ethnicity of the prime was significantly impacted by ethnicity-independent affective cues. Implications for interventions aimed at reducing racial bias are discussed. |
