is the irb getting ready to screw you over again?

I’m actually a fan of IRBs. It is generally a good thing that universities have rules addressing ethical conduct for research. However, IRBs become a burden and nuisance when they overstep their boundaries. While I’ve been lucky in getting my studies approved with little hassle, I have heard from too many colleagues who have had projects delayed with all kinds of petty issues.  The problems seem to come in a few flavors:

  • Treating social science/humanities research like medical research. In medicine, we are often subjecting people to risky treatments in an institutional setting. That’s why you want to get people’s signatures. In social science, we are usually not exposing people to risk in that way and we interact in places where getting a signature would be bizarre (e.g., doing ethnography in a “real” situation).
  • Dropping common sense. For example, it is common to pay people for research participation. Strangely, many IRBs consider payment a form of coercion. Another example is when IRBs ascribe risk to all kinds of innocuous activities. For example, one IRB delayed me because they thought I could use a zip code to track down individual people, even though zip codes usually contain thousands of people.
  • Bureaucracy and delays. IRBs now require approval for just about anything, so they are swamped with everything from major medical trials to freshman oral history projects.
  • Acting as a judge of scientific merit. IRBs are about protecting human subjects, but they sometimes reject projects simply because they don’t like them. It is not the job of the IRB to judge the scientific merit – only risk to human subjects. The stupidest project in the world should be approved without question as long as it conforms with guidelines that protect human subjects.
  • Inconsistency and vagueness. The rules are often vague and hard to follow. Campuses vary widely in their application of the rules. Trying to get multiple campuses to agree on an IRB approval for intermural research can be a nightmare.

In response to these problems, the Office for Human Research Protections, a federal agency, is now trying to revamp the rules. The New York Times reports:

Researchers in the humanities and social sciences are pleased that the reforms would address repeated complaints that medically oriented regulations have choked off research in their fields with irrelevant and cumbersome requirements. But they were dismayed to discover that the desire to protect individuals’ privacy in the genomics age resulted in rules that they say could also restrict access to basic data, like public-opinion polls.

The issue is that the Feds suggested a rule that said data can only be analyzed for the purpose for which it was collected. So, for example, a survey on religion could never be used to study, say, race.

Once again, it seems like the Feds are doing the “one size fits all” approach which will leave the humanities and social sciences in a mess. A federal official is quoted as saying:

Dr. Menikoff said, “We want to hear all these comments.” But he maintained that when the final language is published, critics may find themselves saying, “Wow, this is reasonable stuff.”

Gee, I hope so. The last version left us 5,000 IRBs who can’t agree on anything. What could go wrong?

Adverts: From Black Power/Grad Skool Rulz

Written by fabiorojas

October 26, 2011 at 12:45 am

Posted in academia, fabio

8 Responses

Subscribe to comments with RSS.

  1. I was partly pissed off that ASA didn’t appear anywhere in that article (as compared to anthro, three different historical societies, and the consortium). The only thing I could find about this on was a reference in the April Footnotes to us paying our dues to COSSA (which in turn does good work on the human subjects issue). Assuming that I didn’t simply miss something, I have to wonder why ASA hasn’t done anything directly about one of the biggest problems facing the discipline.

    It would be one thing if ASA deliberately took a minimalist mandate limited to publishing the journals and organizing the annual meeting, but it’s back asswards for ASA to ignore an issue that is core to our professional interests and where we could actually sway administrative lawmaking while taking a position on every hot button political issue that is irrelevant to the practice of sociology and on which we all know the association’s efforts are completely swamped by more powerful political actors. Remind me why exactly we’re paying for them to be on K-Street instead of in (say) Nashville? The way I see it, if the Consortium of Social Science Associations is doing the heavy lifting on lobbying for issues that actually matter to the discipline then I’m all in favor of ASA paying them for their efforts and passing along the costs to the membership. However this also implies that there’s no point in keeping up ASA’s independent lobbying efforts (and the expensive locational decisions that go with that).

    Condo delenda est



    October 26, 2011 at 4:21 am

  2. Re: “dropping common sense”–a zipcode may not uniquely identify someone, but zipcode, sex, and birthdate probably will.



    October 26, 2011 at 5:19 am

  3. When I teach research methods, I have the students do surveys and interviews as part of a class project using an IRB-approved consent form. I find that none of the research participants are remotely concerned about participating until they see the consent form. It is the consent form that feels unethical to them, as it is this document which records their name.



    October 26, 2011 at 5:16 pm

  4. Mikaila: I have a lot of disputes with my IRB (which I’m on, sadly) but one thing they are good on is the provision that signed consent can and should be waived if the consent form itself is the only thing that would link the respondent to the data. That is a pretty standard understanding for many IRBs. You might see if you can get that to fly with yours..



    October 27, 2011 at 12:42 am

  5. I work at a teaching university, which has an IRB but uses a unique approach. Basically, if you’re merely administering a survey, you can submit your proposal and get approval without a committee meeting…so long as it’s written according to their guidelines and the Chair signs off. This is quite helpful for those of us in business.



    October 27, 2011 at 7:30 pm

  6. On the one hand is the complaint that consistency is lacking. (And that is a fair point.) Yet, when the Federal Government wants to act, the worry is that one size cannot fit all. (And that is true, as well.)

    Moreover, professional societies have their own codes of conduct. Understanding the phenomenon of misconduct in scientific research should begin with positives, rather than with negatives. We worry about wrongdoing without defining what it means to do right. Physicists post the Periodic Table in their classrooms and laboratories, but they do not display the “Guidelines for Professional Conduct” of the American Physical Society. Moreover, the APS webpage for that has links to the ethics statements of the American Chemical Society, American Mathematical Society, the Association for Computing Machinery and the Institute for Electrical and Electronic Engineers. Those are all very short statements. The ethics code of the American Counseling Association runs 18 pages. Geographers also have a detailed document.

    So, you can follow one code of ethics, but be blocked by the enforcers of another. Perhaps, it would be better not to have any oversight. But that opens the door to Stanley Milgram and his high school classmate Philip ZImbardo. They say that no one was really hurt. How would they know?

    With human subjects in sociology you risk unleashing unpleasant memories. Is that bad? Is it actionable in some way? In other words, when someone volunteers for a (ahem) “scientific experiment” would they not be responsible for their own participation?

    Maybe the problem with “one size fits all” is that one answer does not resolve all aspects of the same question. In other words, it’s complicated; and it cannot – should not – be simplified into rules.


    Michael Marotta

    October 27, 2011 at 9:34 pm

  7. I am generally a pretty strong critic of IRB overreach.

    However, there is an important point to be made about IRBs as arbiters of scientific merit in certain situations.

    Evaluating a proposal with (real) risks to participants involves weighing the risks with the potential benefits (to participants, to research, and to society). The quality of the study design bears on the benefit side of that equation.

    So, in cases (like medical research IRBs) where there is the potential for serious risks to participants, scientific merit should probably be considered.



    October 30, 2011 at 10:00 pm

  8. […] orgtheory, we’ve discussed how researchers face challenges concerning the IRB here and here. Although the IRB offers detailed guidelines that can protect human subjects in medical […]


Comments are closed.

%d bloggers like this: