Primed and Prejudiced: Investigation of the Generalization of Black-violence Stereotypes to Biracial Faces in the Weapons Identification Task

Elizabeth Esser-Adomako

Advisor: Matthew S. Peterson, PhD, Department of Psychology

Committee Members: Craig McDonald, Martin Wiener

Online Location, Online
August 12, 2022, 01:00 PM to 03:00 PM


The rapidly increasing Multiracial population in America has inspired a new body of research devoted to understanding perceptions of racially ambiguous faces. This work has found that Black-White Biracial faces are more frequently categorized as “Black” than “White”, broadly supporting the existence of hypodescent, or the tendency to classify Multiracial individuals according to their lower status group. However, more recent research also demonstrates that hypodescent, is less prevalent when a “Multiracial” categorization option is available. Critically, little is known about the extent to which biases toward monoracial “parent” groups generalize to Multiracial individuals. The results across five experiments demonstrate trait hypodescent, in which Black-violence bias generalizes to Black-White Biracial faces in a sequential priming task. Experiments 1 and 2 included real and morphed Biracial faces, respectively, along with Black and White face primes in the Weapons Identification Task (WIT; Payne, 2001); Experiments 3, 4, and 5 included real and morphed Biracial faces within the same task. The results across these experiments suggest that Black-violence biases extend more readily to morphed than to real Biracial faces in the WIT. Results of a face categorization task robustly demonstrated that morphed Biracial faces are more frequently categorized as “Biracial” and “Black” and less frequently categorized as “White” and “Other” compared to real Biracial faces. The combination of WIT and categorization results support the theory that the morphed Biracial faces are perceived and categorized as Biracial and Black more often. Implications of the differences in results between morphed and real Biracial faces are discussed further, including the need for greater variability in the images used to create morphs and an examination of which features most strongly cue race. The current findings also indicate that internal and external motivations to respond without bias (IMS and EMS; Plant & Devine, 1998) play an important role in the degree of trait and face hypodescent seen in response to Biracial faces in the WIT and face categorization task.