Purpose: Measures of optimistic bias are used to examine whether a group is unrealistic in their estimates for experiencing a negative health outcome. Few studies have examined differences between direct and indirect approaches to assessing optimistic. This study uses both approaches to measure optimistic bias for HIV infection, and examines the relationship between optimistic bias and HIV risk behaviors.
Method: Participants were recruited from HIV/STD testing and risk-reduction counseling programs in Long Beach, California. Participants (N=455) completed a direct measure (e.g. “compared to an average other, do you think your chances of getting HIV are: less/about the same/greater”) and two items indirectly assessing optimistic bias (e.g. separate items assessing risk of HIV infection for the self and an average other). If mean responses to these measures are no different from the midpoint, than this suggests the sample is not biased in their perceptions. Participants completed the Risk Behavior Assessment which includes items assessing demographics, sexual-risk and drug-use behaviors.
Result: In response to the direct measure of optimistic bias, 23.2% of participants self-reported that their chances of getting HIV was greater than the average other, 23.2% that their risk was about the same, and 53.6% that their risk was less than the average other. The sample demonstrated optimistic bias on the indirect (m=1.35, t(438)=11.04, p<.001) and direct measures (m=-.82, t(443)=-9.59, p<.001). Responses to the direct item were correlated with number of sex partners (r(441)=.25, p<.001), and having unprotected oral sex (r(443)=.21, p<.001), unprotected receptive anal sex (r(441)=.14, p<.01), unprotected insertive anal sex (r(442)=.18, p<.001), and having used amphetamines (r(441)=.11, p<.05). Correlations with the indirect measure were smaller, ranging from r(438)=-.07, ns, to r(431)=-.12, p<.05.
Conclusion: Participants recruited from HIV testing and risk-reduction counseling programs demonstrated optimistic bias towards their risk of HIV infection on direct and indirect measures. Participants engaging in more risky behaviors were less optimistic about their risk of HIV infection. The associations between HIV risk behaviors and optimistic bias were stronger when the direct approach to measurement was used.
Candidate for the Lee B. Lusted Student Prize Competition