In early February, Newsweek reported on Jason Chan and Anindya Ghose’s study that claimed ads for casual sexual partners on Craigslist.org resulted in large numbers of new HIV infections. When the mainstream press covers a research study about hot button issues, a good study can have unintended negative consequence, and a bad study can become the basis of policy. This study has serious problems in design, assumptions and in its citations that reduce its credibility. Gay/bi and other men who have sex with men became invisible in this study of the Internet and HIV transmission.

The Management Information Systems Quarterly published this epidemiological study. Sometimes an outside perspective can uncover something new. Other times an outside perspective can just be strange, and not in the good sense.

HIV does not spread randomly in the population like the flu. It spreads in certain social groups and not in others. Gay/bisexual men accounted for 53.2 percent of all new positive HIV tests in 2012, exceeding all other groups. Yet these researchers managed to discuss increases in HIV infections without once using the words “gay,” ‘bisexual,” “man who has sex with men” or “MSM” in the text. The number and content of ads for Man-Seeking-Men on Craigslist should be critical variables for this study to examine, but they were not.

On page 7 of their study, the authors describe the problem driving increases in HIV infections as a choice between having casual sex and enjoying “STD/HIV free status via abstinence.” This false dichotomy ignores the increasing number of prevention choices that people have.

Chan and Gross charge that HIV infected men either withhold or misrepresent their HIV status to on-line sex partners, thereby driving new infections. This scapegoating of the Aware Infected ignores the dangers of the Unaware Infected. They can be either people who have never tested or people who have become HIV-infected after their last negative HIV test, but have not yet re-tested for HIV. The latter can honestly disclose an incorrect HIV status.

The authors cite two studies to support their claim of HIV positive culpability. Only one study (Moskowitz and Seal 2010) has been posted on-line. This study did not focus on HIV itself, but a psychological theory in the context of HIV. Moskowitz surveyed men in multiple countries who had placed internet ads seeking casual male sexual partners, and had a low response rate (10-15 percent). The low response rate and the global reach make it difficult to generalize from these results to trends in the US. Men with HIV infection did report less disclosure and less condom use than did men without, but Moskowitz defined “disclosure” as discussing HIV status with a partner. Moskowitz did not report whether the respondents considered information posted in the ad as part of a “discussion.”

Chan and Ghose also cite McFarlane et al 2000 to support their statement “Irresponsible and risky behavior of having sexual contact with online partners prior to HIV testing is also observed.” McFarlane collected the data as part of a HIV test. The study collected data on behavior in the six months prior to the HIV test. As a result, all encounters would be prior to this this HIV test. McFarlane did not report how many prior HIV tests each subject had, or if this was their first HIV test.

These problems mar this study and its truncated Newsweek “incarnation.” In the tradition of HIV as “punishment for transgressing sexual norms,” this study fails to clarify how the Internet affects HIV transmission.

To read the Chan and Ghose article visit http://papers.ssrn.com/sol3/papers.cfm?abstract-id=2035585