Abstract
Study 1 examined the perceived association of AIDS and death by showing that thinking about AIDS increased participants'; death-thought accessibility. Hypotheses about the consequences of this association for perceptions of people with AIDS were derived from terror management theory, which proposes that mortality salience increases derogation of those who threaten people's worldviews unless those worldviews oppose prejudice, in which case mortality salience can increase acceptance of people who are otherwise threatening. Consistent with these hypotheses, conservative participants had less favorable impressions (Study 2) and liberal participants had more favorable impressions (Study 3) of a target with AIDS following a death reminder. Study 4 suggested that the decrease in prejudice among liberals following mortality salience was a genuine decrease in prejudice (as indicated by responses to an unobtrusive attitude measure), not just an increase in the desire to appear nonprejudiced. Theoretical and practical implications of these findings are discussed.
ACKNOWLEDGMENTS
This material is based on work supported under a National Science Foundation Graduate Research Fellowship. We thank Alexander Engelman, Elizabeth Hooker, Emily Major, and Leschia McElhaney for their data collection and entry efforts.
Notes
1Although this questionnaire was intended as an individual difference measure, we included it after the social distance measure so that participants would not be primed to control their prejudice when evaluating Dan. An analysis of variance indicated that the experimental manipulation (i.e., the death reminder) did not affect responses to the internal subscale, F(1, 74) = .20, p = .66, and only marginally affected responses to the external subscale, F(1, 74) = 3.27, p = .08, so the positioning of the measure in the study does not appear to have affected its ability to independently predict motivation to control prejudice.
Note. Numbers in parentheses are standard deviations. The political orientation scale ranged from 1 (very conservative) to 9 (very liberal). All other scales ranged from 1 to 5.