Ethnography and the IRB

A perennial favorite: the Institutional Review Board and the practice of ethnographic research. I am a member of the Institutional Review Board at my University, ostensibly representing the social and behavioral sciences. Recently the board has undergone a change of leadership and staffing, which means new people and a new round of education about how anthropologists conduct ethnographic research as well as how non-anthropologists do so. For those who don’t know, Institutional Review Boards are a result of the 1974 the National Research Act (Pub. L. 93-348) , which commissioned a report on the use of Human Subjects in Research (known as the Belmont Report) and upon which a code of federal regulations 45 CFR 46 is based, mandating that all research institutions receiving any federal support must oversee research using human subjects. If it isn’t research, or they ain’t human, the IRB is not concerned–otherwise, they must make every effort to 1) know about it and 2) approve it. So far so good.

There are however, a couple of sticking points with respect to ethnnographic research that I have run into repeatedly. One is the requirement of confidentiality. The other is the requirement of written informed consent (and here). (There is also the problem of whether or not there are any reliable standards when it comes to “psychological harm” that subjects might suffer at the hands of unscrupulous researchers–but that is a bugaboo for another post). Now, given the anguish Anthropology has put itself through with respect to colonialism, with respect to complicity in the dispossesion of native people’s land, and most recently with respect Napoleon Chagnon and Patrick Tierney’s Darkness in El Dorado, one would probably be safe in asserting that we are alone among the disciplines who have made the ethical treatment of humans not just a priorty, but virtually impossible to ignore throughout one’s graduate and professional life. We may not have standards of harm that are easily applied across all cases–but we damn well have discussed, in excruciating detail, nearly every possible kind of injustice man can visit on man.

My new IRB overlords, whom I welcome of course, do not see it this way. As far as they are concerned, anthropology appears to occupy a backwater, in which the lack of standardization, the reliance on individual judgment, and the practice of meticulously examining the finest grain of every encounter we have for its justice or injustice are nothing so much as solid evidence that we are in violation of what are clear and reasonable regulations about how scientific research should be conducted.

Take for example, the requirement of written consent. August authorities such as the Anthropological Association of America and the former head (Project Muse required) of the National Science Foundation Anthropology Section have made clear that written documentation may not be appropriate in all cases–the easy cases are those where your informants are non-literate (duh?) or where you are dealing with individuals who have, for instance, a long history of land expropriation due to the unscrupulous signing of documents by ancestors. In these cases they recommend that some other method of obtaining the vaunted “informed consent” of research participants be pursued. But there are a host of less easy cases: those situations where your informants are naturally or historically suspicious of government and its apparatus for instance, or where the presentation of a form would convert a personable and arduously cultivated relationship into a purely formal one. Should the IRB require the presentation of such forms in all cases–even to the extent of rendering some forms of long-practiced ethnographic research impossible? My new IRB Overlords appear to think so.

Or take for instance, the issue of confidentiality. According to the IRB, the only ethical practice is to ensure confidentiality of information, as a default practice, and if necessary, against the wishes of informants. Sure, if the informant is a public figure and the information is public, and the informant wishes to have it made public, then in that case, there seems to be no problem with allowing the researcher to treat the information as public. But if, for instance, an individual of no particular standing, in negotiating whether to participate in your research project, demands that his words be made public and be identified with his real name–do we have the right, or the obligation, to deny him this request? It seems to me that applying this ethical standard across all cases, regardless of the express wishes of individual informants is as suspect as revealing their confidential information. It is, as we should well know, the moral ideal of anglo-american privacy elevated to the status of universal norm. My new IRB overlords call BS on that.

So what I want to know is: experience anyone? I need examples of research in which either the use of oral/non-written documentation of informed consent, or the practice of allowing people the option to be identified has been approved (recently) by IRBs, or articles and documents that describe, or defend such practices. Ultimately, I would like to pen recommendations for use in every such changing of the guard… because otherwise, institution after institution is set to make the practice of ethnographic research as we know it both increasingly formalized and possibly, increasingly impossible.

Christopher M. Kelty is an associate professor at the University of California, Los Angeles. He has a joint appointment in the Institute for Society and Genetics, the department of Information Studies and the Department of Anthropology. His research focuses on the cultural significance of information technology, especially in science and engineering. He is the author most recently of Two Bits: The Cultural Significance of Free Software (Duke University Press, 2008), as well as numerous articles on open source and free software, including its impact on education, nanotechnology, the life sciences, and issues of peer review and research process in the sciences and in the humanities.

18 thoughts on “Ethnography and the IRB

  1. Here’s the final sentence in an almost page-long confidentiality agreement for an NIH study: “The investigators may also voluntarily disclose information, for example, if it is required by the Food and Drug Administration (FDA), members of Congress, and law enforcement officials”.

    So just tell the folks who want their real names used to get a Tom DeLay type to demand it. (Hey, maybe they could hire Jack Abramoff to fix it!)

  2. At the last AAA meeting, there was a panel called “The Ethnography of Academic Life”, which had a paper about the IRB that you would have found interesting (assuming you weren’t there) because the presenter seemed to be dealing with rather similar issues. I took some notes on it:

    Hyper-vigilance at the federal level has affected the IRB process. Regulations that were designed for biomedical research are being applied to the social sciences. She uses “non-violent interpretations” in her ethnographic study of history, sociology, and journalism in the IRB process. Research in other disciplines may seem suspect to anthropologists, and anthropological research may seem suspect to other social scientists. She was a member of an IRB. Other disciplines on the IRB panel saw ethnography as shady because they aren’t using the “scientific” model of dependent and independent variables, etc. When she encountered the word “debriefing”, the word was unfamiliar to her but very familiar to the social scientists in the other disciplines. Debriefing means explaining to research subjects how and why they were deceived during the research process. Deception is not an accepted part of anthropological ethics. Oral historians are horrified by making transcripts and recordings anonymous because in their discipline, the goal is to make specific, attributable historical records available for future reference.

    Also, here are some articles which may or may not be useful:

    Anthropology News
    2003 Oral History Excluded from IRB Review. Anthropology News 44(8):29.

    Association, American Anthropological
    2004 American Anthropological Association Statement on Ethnography and Institutional Review Boards: Electronic document,, accessed February 9, 2006.

    Fitzgerald, Maureen H.
    2005 The Ethics-Review Process. Anthropology News 46(6):10-11.

    Gray, Mary L.
    2003 The Plasticity of Vulnerability: Research with a Stigmatized Community. Anthropology News 44(8):19.

    McGouch, Helen
    2004 IRBs and Ethnographers: Why Can’t We All Just Get Along? Anthropology News 45(2):7.

    Plattner, Stuart
    2004 Human Subjects Protections and Anthropology. Anthropology News 45(2):5-6.

  3. I’m a linguist but we have the same problems with anonymity and informed consent. I work in Australia and so far I have received exemptions under Title 45 CFP 46, in part because I follow guidlines for research set down by the Australian Institute for Aboriginal and Torres Strait Islander Studies, and the Land Rights Act (1978, if memory serves). The Canadian SSRC has a set of guidelines which I have also referred IRBs to in the past, and to the AAA’s code of ethics.

  4. Me again, sorry forgot to add (since this might be of general interest and since no one’s brought it up yet) there are provisions for not getting written informed consent. It’s in the legislation, and I can look it up if you can’t find it. The upshot, at least as far as I know, is that if informed consent is deemed necessary, it must be in writing, but if the research is exempted from a full review (as linguistic fieldwork tends to be) the legislation is not quite a problematic.

  5. My experience was that written informed consent could be waived at the IRB’s discretion, assuming that I could reasonably claim adverse circumstances as described elsewhere here – nonliterate population, extreme mistrust, etc. In place of a form, I was told to create a script for obtaining a verbal waiver of signed consent. This was not, I discovered, a set of notes for how I might explain what I was doing to a potential research participant – it was a literal word-for-word script that I would be required to read to each participant in order to obtain said verbal waiver of signed consent. My script went back for several rewrites, including corrections of my grammar in Spanish by the IRB officer. Let me just say that in practice, getting a research participant to sit still through the 7-minute reading of my script proved challenging.

  6. These are great (thanks Noah, Claire, and Adam!), but unfortunately not re-assuring. Adam’s experience is precisely the kind of bloody-mindedness that I am trying to prevent here– because in practice its never going to happen, and if the only point is bureaucratic intimidation, well….

  7. I doubt it is bureaucratic intimidation — more likely covering their behinds and wanting to be on the right side of history (no-one wants to be the review board that approves a modern Tuskegee experiment). So I would imagine the most useful strategy would be to talk about how the work of the IRB is ultimately about substance, not process — are people appropriately protected? How do we guarantee the protection of human subjects in a thoughtful rather than a formulaic manner? The 7 minute document Adam mentions having been forced to read by his IRB board could actually be a *great* illustrating anecdote for you Chris: clearly it didn’t do any research subjects any good at all or protect them from anything; it sounds rather more like those statements read out by the Spanish in the Americas before undertaking conquests. Making that kind of point could get your new colleagues on your side — I would guess their main motivator is the fear of failing at an important task.

  8. Ozma, in my optimistic moments I totally agree… I’d love to believe that my Overlords are good-natured and well-meaning and we can find ways to actually protect people. The rest of the time, I worry that they are do-gooders with a very narrow idea of what it means to protect people, and a deep, irreconciliable suspicion of any kind of research that does not follow those ideas. I think they are motivated by fear of failing–but I also think they are motivated by other, more insidious things, like following what they perceive to be “best practices” in the IRB world, which are “best” only because lots of people follow them; or by fear of lawsuits.

  9. well, that’s probably true — but the more it’s true the more you (and by “you” I mean “the practice of ethnography by anthropologists”) are screwed. So what can you do but give the optimistic approach the old college try?

  10. I think one of the main problems is that in some cases the members of HSIRB boards do not see ethnography as a valid form of research. For my master’s thesis, at my current institution, I was able to go through the expedited review process because I met certain criteria under the various laws that govern HSIRB. However, I have heard horror stories of people who have to go through the full review process and submit their HSIRB proposals to the full board. In many cases, this resulted in critiques of ethnography as a methodology and in some cases theoretical ideas (e.g. Asking one student how they would “measure” the homosexuality of informants.) There seems to be a strong assumption among HSIRB boards that ethnography is not objective and/or that subjective research is in of itself a bad thing. This seems to lead some HSIRB boards to assume that since enthnography is “flawed” because it is subjective then this methodology could lead to harm to subjects. This leads to the unfortunate situation where some HSIRB boards end up critiquing the entire basis for much anthropological methodology and theory instead of actually looking at the specific, unique circumstances of an individual research project.

  11. IRBs vary greatly and there is at present no realistic “check” on their power. The AAA guidelines are not particularly helpful, either; they are vaguely worded and so broad that they are generally unhelpful when an anthropologist is trying to educate an uninformed Board. I’ve been a member of an IRB, I’ve headed an IRB and I now fund programs that (of course) require IRB permissions, and when dealing with their bureaucratic requirements, one must realize that there is legally actionable consequences that the university will be liable for should they approve a request that turns out to be less than ethically pure. Good, clear guidelines from the AAA that are more up to date than the current ones would help these committees (and the researchers, and ANTHROPOLOGY) beyond measure. Anyone try to do post-911 research on Islamism? Anyone successful at getting IRB approval to interview failed suicide bombers in foreign locales? What about approvals for studying IRC/web-chat, blog participation, forums? IRB protections in such instances are completely in a muddle. The law is not going to catch up with these problems anytime soon. It’s high time that AAA updated its guidelines to effectively assist researchers to determine what is and is not ethical in the modern world.

  12. I conducted my dissertation field research among responders and security planners in a U.S. city from June 2001 through December 2003. The initial IRB issues (“and this is how I plan to reduce risk for the nice men with guns….”) were strange enough. After 9/11 when I needed to renew, I just had to cross my fingers and hope I had given them enough butt covering material to let me keep going. Although my research subjects were not vulnerable in the traditional sense, political and security concerns made them reluctant to sign informed consent forms. Practical concerns made it impossible to use the verbal consent method.

    I am willing to share my IRB applications, discusss the revisions, and how I handled it in the field. My contact info is on my website.

  13. I know it’s a different country with different legislation and all, but the same ass backward processes underwritten by concerns about liability neo-liberal audit that have nothing to do with ethics…or people for that matter. Anyway… here at the University of Toronto Gavin Smith is now chair of our ethics review board and he’s come up with some ways of incorporating participant observation into the ‘traditional’ ethics protocol. I’m in the middle of the process myself and have am finding this stuff particularly useful. You can check out the details at
    Also, one of the ways that folks in the dept have gotten around the written consent thing before is by arguing that it is culturally inappropriate, which may or may not work for you depending on who you do your FW with.

    anyone have an informed consent script they’d be willing to share? Mine keep coming out like total legalese or like they were written for, say, the target audiance for teletubbies.

Comments are closed.