INVESTIGATING RACIAL BIAS IN PERCEPTIONS OF FREE WILL

The overarching goal of this study was to examine whether perceptions of others’ free will would differ depending on perceiver race as well as target race. The current study proposed that such a racial bias may be one mechanism by which racial disparities in medical treatment recommendations arise. By bridging findings from four different lines of research (i.e., the literatures on racial health and medical treatment disparities, racial bias, free will beliefs, and social identity), it was hypothesized that: (1) participants would perceive greater amounts of free will for a hypothetical racial ingroup patient than an outgroup patient; (2) such effect would be moderated by participant racial identity and/or racial bias, such that greater racial identity and/or ingroup racial bias would result in greater differences in racial ingroup vs. racial outgroup members’ free will; and (3) greater perception of the patient free will would indirectly affect treatment recommendation for the patient through increased perceived patient self-control. In order to test these hypotheses, the study used a 2 (Participant race: Black vs. White) x 2 (Target race: Black vs. White) x Continuous (Racial Identity/Racial Attitudes) between-subjects design, in which target race was manipulated experimentally. The results indicated that Black participants’ perceptions of patient free will was moderated by both racial identity and racial bias. Specifically, those who weakly identified with their racial group perceived a greater amount of free will in the White target patient than the Black target patient. Also, Black participants who displayed pro-White racial bias, a greater amount of free in the White target patient than the Black target patient. These moderating effects of racial identity/racial bias were not found for the White participants. Also, patient free will had an indirect effect on treatment recommendation by way of perceived patient self-control, such that perceived free positively predicted the more rigorous of two treatments. Limitations of the current research include the undergraduate college student sample, the use of a general measure of racial identity, and the use of the old IAT algorithm. Future work should examine empirically whether findings from the present study can be generalized to provider samples