Tuesday, August 21, 2007

Northwestern IRB: Unsystematic Interviews Are Not Subject to Review

Today's New York Times features a story, "Criticism of a Gender Theory, and a Scientist Under Siege," about the case of J. Michael Bailey, Professor of Psychology, Northwestern University. Bailey's controversial book about identity. The book provoked several complaints, including the charge by "four of the transgender women who spoke to Dr. Bailey during his reporting for the book . . . that they had been used as research subjects without having given, or been asked to sign, written consent."

As reported by the Times, the case was investigated by Alice Domurat Dreger, Associate Professor of Clinical Medical Humanities & Bioethics at Northwestern, who has posted a draft article on the subject, "The Controversy Surrounding The Man Who Would Be Queen: A Case History of the Politics of Science, Identity, and Sex in the Internet Age," [PDF]

Dreger finds that Bailey did not commit serious ethical violations, nor did he violate the requirements for IRB review:

the kind of research that is subject to IRB oversight is significantly more limited than the regulatory definition of “human subject” implies. What is critical to understand here is that, in the federal regulations regarding human subjects research, research is defined very specifically as “a systematic investigation, including research development, testing and evaluation, designed to develop or contribute to generalizable knowledge” (United States Department of Health and Human Services, 2005, sect. 46.102, def. “b”). In other words, only research that is truly scientific in nature—that which is systematic and generalizable—is meant to be overseen by IRBs. Thus, a person might fit the U.S. federal definition of “human subject” in being a person from whom a researcher gains knowledge through interpersonal interaction, but if the way that the the knowledge she or he intends to gain is unlikely to be generalizable in the scientific sense, the research does not fall under the purview of the researcher’s IRB.

It is worth noting here, for purposes of illustration of what does and doesn’t count as IRB-qualified work, that I consulted with the Northwestern IRB to confirm that the interviews I have conducted for this particular project do not fall under the purview of Northwestern’s IRB. Although I have intentionally obtained data through interpersonal interaction, the interview work I have conducted for this historical project has been neither scientifically systematic nor generalizable. That is, I have not asked each subject a list of standardized questions—indeed, I typically enjoyed highly interactive conversations during interviews; I have not interviewed all of my subjects in the same way; I have negotiated with some of them to what extent I would protect their identities. This is a scholarly study, but not a systematic one in the scientific sense. Nor will the knowledge produced from this scholarly history be generalizable in the scientific sense. No one will be able to use this work to reasonably make any broad claims about transsexual women, sex researchers, or any other group.

When I put my methodology to the Northwestern IRB, the IRB agreed with me that my work on this project is not IRB-qualified, i.e., that, although I have obtained data from living persons via interactions with them, what I am doing here is neither systematic nor generalizable in the scientific sense.

Clearly Bailey's work hurt the feelings of some people he wrote about, but, as Dreger notes, "scholarship (like journalism) would come to a screeching halt if scholars were only ever able to write about people exactly according to how they wish to be portrayed." Indeed, that's what social scientists have been arguing for three decades.

No comments: