Abstract
Aims To investigate the implicit (unconscious) and explicit (conscious) attitudes of primary care providers (PCPs) toward African Americans and Latinos, and compare them to the attitudes of community members.
Methods A computer-based survey was completed by 210 PCPs (60% participation) and 190 community members (55% participation) in primary care clinics of three different health care organizations throughout the Denver metro area. Results were analyzed in aggregate. The survey measured attitudes toward African Americans and Latinos, compared with Whites in each case. Implicit attitudes were measured with Implicit Association Tests (IAT) and explicit attitudes were measured by self-reported scale ratings.
Results A wide range of implicit attitudes was found among both PCPs and community members, from greater favorability toward Latinos or African Americans to greater favorability toward Whites (Latino:White IAT range = −0.91 (pro-Latino) to 1.28 (pro-White); Black:White IAT range = −1.11 (pro-Black) to 1.32 (pro-White). Consistent with other research, implicit pro-White bias was the more frequent and stronger response (Latino:White IAT M=0.31, Cohen’s d=0.78; Black:White IAT M=0.27, Cohen’s d=0.76), and the two samples did not differ in their implicit bias (p > .20). The only consistent demographic predictor of implicit bias was participants’ self-identified ethnicity/race. White participants demonstrated stronger pro-White bias than did minority participants. A comparison adjusting for differences in the ethnic/racial composition of the PCP and community samples showed that once these differences were controlled, PCPs showed less pro-White bias than the community members. Both groups reported little explicit bias (Cohen’s d=0.04), with most participants reporting they felt similarly toward the ethnic/racial groups.
Conclusions Numerous studies have shown that implicit ethnic/racial attitudes in the U.S. generally favor Whites over other ethnic/racial groups. The current study finds that implicit attitudes among PCPs and community members in the Denver metro area are no exception. Despite these general trends, this study also shows that people are not all the same, with some individuals demonstrating strong pro-Latino or pro-African American bias while others show strong pro-White bias. Phase two of the study will begin to investigate whether such biases are related to patient care and hypertension control.
