About Me

Find out more about me here.

05 November 2012

IRB Comments on Methodology

The primary purpose of the human subjects committee is to protect the people whom we are studying. It is that simple. (While it is not our intention and should certainly not be our concern, by protecting subjects, we also shield the university from litigation.) Among the things that we need to ensure are the following: all human subjects who are part of research done at or in conjunction with a given institution must consent to participate, must have the ability to refuse to participate and to withdraw their participation, can expect that they will not be exposed to unreasonable risks, and can expect that their identity will be protected as either anonymous or confidential.

Beyond that, it is not the IRB's job to ensure that the research that we review is good research. It should be the policy of IRB's that we should not critique our colleagues' methods. While it can sometimes be tempting to do so, it is problematic for several reasons. First, the IRB is a black box so any animosity that a researcher harbors for her reviewer gets transferred to the board. We do ourselves--and ultimately would-be human subjects--no favors by garnering the contempt of our peers. IRB's are despised enough without us further fanning the flames. Second, we are busy people with competing demands; unsolicited methodological critique should not be another. Finally, we have a duty to be good neighbors. We must respect the training, intelligence, and expertise of our colleagues, and in the case of student research, we must respect the abilities of the student's faculty adviser. Unless you are asked directly for your scholarly opinion, do not give it. Peer reviewers for conferences, university presses, and academic journals as well as annual faculty evaluation committees, department chairs, and tenure and promotion committees are more appropriate bodies for the review of faculty research.

That said, there are some circumstances under which poor research design may spill over into our commission to oversee the ethics of that research. For example, imagine a protocol that proposes to stick subjects with needles to gauge their favorite color. This research design is flawed since it would not be testing what it claimed to be, but since the subjects are being exposed to pain and an increased risk of infection and because the research design is flawed, the pain and risk would be pointless. In other words, the study could be conducted in a less invasive manner with equal or better results. However, not all methodological errors are equally dangerous. A question on a survey that asks, "What is your favorite color? (a) blue (b) bright (c) dull (d) green," might be poorly written but, because it does not expose the participant to any additional risk, it falls outside of our purview.

No comments:

Post a Comment