Can a Short Film Impact HIV-Related Risk and Stigma Perceptions? Results from an Experiment in Abuja, Nigeria

Abstract

HIV/AIDS-related stigma is believed to result in negative social consequences for people with the disease and to be a deterrent to HIV serostatus testing. The ability of communicators to change people's stigma perceptions and subsequently impact decisions to test, however, is not well understood. Based on the entertainment–education approach, this article presents the results of a field experiment conducted in Abuja, Nigeria, testing a mediated intervention designed to reduce HIV-related stigma and risk perceptions. The results indicate that the intervention was effective relative to a control in impacting perceptions of the severity of HIV and some stigma-related attitudes, particularly for male participants; and that for this sample, risk and stigma perceptions significantly impact intentions to test for HIV. It also showed that severity perceptions mediated the relationship between viewing the film and testing intent.

ACKNOWLEDGMENTS

The authors would like to thank the staff at the Center for the Right to Health for their assistance with data collection.

Notes

¹As a check on random assignment in the final experiment, we conducted analyses to determine whether or not age of participants and sex of participants was evenly distributed across the film and control conditions. We found that for those participants who reported age, age of participants did not differ across conditions, t(85) = 1.56, p = .12; r =.05. There were more men than women in the sample, but cross-tabulation of the data indicated that these groups were not unevenly distributed across conditions, χ ² (1) = .011, p = .98. Further, people who self-reported testing for HIV in the past did not differ across conditions, χ ² (1) =3.15; p = .09. Given these findings, tests of hypotheses and research questions were conducted.

²Given the differences in cell size for male and female participants by film condition (see Table 2), and the sensitivity of analysis of variance to violations of assumptions of unequal cell size, t tests were conducted for these analyses for males and females separately rather than using analysis of variance to test the predictions.

³The order of entry of the blocks was chosen based on Tabachnick and Fidel's (2001) recommendation that variables be entered into the regression equation based on theoretical importance, beginning with the most important. Entering the stigma variables prior to entry of the risk perception variables did not substantially alter the results. Inclusion of education as an independent variable in the regression did not substantially alter the results.

⁴Given that the film had a consistent effect on men but not women, we tested this model for men only. The experiment did not influence the same stigma variables that were related to testing behaviors; thus, these factors were not included in the model and a separate model could not be tested for women.