“I'm Not Telling”: The Effects of Racial Disparities Communications on Task Persistence Among Blacks

Abstract

Research suggests that communications about racial health disparities may adversely affect Blacks. In this study, we varied the message content (Black–White cardiovascular-related disparities + neutral health topics vs. neutral health topics only) embedded in public service announcements given to Black and White participants (N = 86) and had them complete a purported health self-assessment. We used the number of items completed as a measure of task persistence. Our results showed that participants in the disparities condition completed fewer items on average than participants in the neutral condition (p < .01). Planned contrasts revealed that this effect was driven by the responses of Blacks who completed fewer items in the disparities condition (p < .01), though Whites evinced a comparable condition-based trend (p = .12). We found no Black–White differences in the number of items completed in either of our experimental conditions (ps ≥ .53). Although preliminary, our findings suggest that Blacks and Whites exposed to comparative racial disparities messaging about cardiovascular diseases could experience reduced task persistence. Research implications and study limitations are also discussed.

Acknowledgment

All methods and procedures were approved by the institutional review board at the Pacific Institute for Research and Evaluation.

Notes

¹We equate the term White with non-Hispanic White and the term disparities with differences and inequalities in this article.

²Nicholson and colleagues’ (2008) Frame × Medical Mistrust interaction was marginally significant (p = .07) and only reached significance when they excluded data from the impact condition.

³Two participants did not indicate their age. In addition, 18 participants requested and were granted an additional $10 to cover their travel. Because our findings remained consistent when these participants were included or excluded, we retained these data in our statistical tests.

⁴One participant did not provide nativity information.

⁵Beyond a marginally significant experimenter effect on the PSA clarity item (p = .077), we found no reliable effects involving this variable on our measures of interest (ps ≥ .17). We therefore collapsed our data across this variable in our analyses.

⁶Although no participant requested his or her LPI score, several participants asked about their health status at the end of the study. Therefore, we explained the reasons why participants would not be receiving this information in the debriefing.

⁷The term provocative refers to items we identified as either offensive, highly personal, or private in nature and therefore capable of eliciting a response independent of our manipulation. For instance, the Carroll Rating Scale for Depression item “I often wish I were dead” (Carroll et al., 1981) was removed from the LPI because of our concerns that it may cause participants to feel uncomfortable and exit the survey.

⁸We tested this indirect effect with a 95% bias-corrected bootstrap confidence interval based on 1,000 bootstrap samples.