CITI is still misrepresenting Milgram’s obedience research


Two years ago, after taking the Collaborative Institutional Training Initiative (CITI) online ethics training course required by my institution, I wrote to them to object to the way Milgram’s obedience research was characterized in their “History and Ethics” module. Short version: they compared Milgram’s research to Nazi medical experiments and the Tuskegee syphilis study.

In response to my email, I got a very nice sounding reply from a CITI staff member. Quoting from her email:

I agree with you.

The module was adapted from a module written for biomedical researchers. When it was adopted, in order to make it more relevant for researchers in the social and behavioral sciences, the writers simply added cases that seemed more relevant. The important distinctions you note were not made.

I would like to revise the module completely and the only obstacle right now is time.  I will see if I can get some minor changes approved, in the meantime, that will address your issues. One simple solution might be to change the introductory language to the case studies and remove the word scandals where it is not appropriate.

But yesterday when I took my requried CITI refresher course, I discovered that the promise was an empty one. Not a word has been changed.

So I’m back to writing to them. Below is my latest email. Below it, I have quoted the objectionable text from the CITI course.

Dear CITI staff:

Two years ago, I wrote to you to object to the way that Stanley Milgram’s obedience research, and other behavioral science research, was portrayed in your “History and Ethics” module. That email is appended below.

The crux of my objection was that CITI mischaracterized Milgram’s research as unethical and drew parallels to Nazi medical experiments and the Tuskegee syphilis study. To the contrary, Milgram’s obedience research was conducted ethically (in fact, it was replicated with IRB approval just a few years ago). It is indeed relevant to contemporary research ethics — not as an exemplar of harmful research, but because of what it teaches us about how research subjects may respond to scientific and institutional authority.

In response, I received a message from Lorna Hicks (appended below) in which she stated that Paul Bruanschweiger had forwarded her my email. She stated quite bluntly, “I agree with you.” She assured me that CITI would update its materials. At that time, I was pleased both with CITI’s prompt responsiveness to feedback as well as with the specific substance of the reply.

So perhaps you can imagine my surprise and dismay when I sat down to take the CITI refresher yesterday — two years later — and discovered that Milgram and several other behavioral studies are still being described as “similar events” to Nazi war crimes and the Tuskegee syphilis study. Despite your assurances made two years ago, the module has not been changed to remove the objectionable comparisons.

So once again, I am writing to you to strongly object to your characterization of Milgram’s obedience research. You are doing a disservice to the legacy of an important body of behavioral science research, and you have continued to do so for several years despite agreeing that it was wrong and promising to stop.

Sanjay Srivastava

Here is the text of the CITI “History and Ethics” module that I objected to, as it appeared both in 2009 and 2011. I have quoted at length to provide the full context, so you can see the comparison for yourself. Boldface emphases have been added by me.

The development of the regulations to protect human subjects were driven by scandals in both biomedical and the social/behavioral research, and as such reflect social concerns regarding research involving human subjects including:

* The importance of meeting the requirements of basic ethical principles underlying the involvement of humans as research subjects

* The need for independent, objective review of research

* The need to preserve the public trust in research involving human subjects

1.0 Historical Development

The events that led up to the development of the currently regulatory system occurred in both biomedical and social/behavioral research.

1.1 Events in Biomedical Research

Attention to the ethics of human subjects research first received wide-spread attention after WWII with revelations of the Nazi “research” which led to the Nuremburg Code, a statement of ethical principles of human experimentation. In 1964, the World Medical Association developed a code of research ethics that came to be known as the Declaration of Helsinki. It was a reinterpretation of the Nuremberg Code, with an eye to medical research with therapeutic intent. In 1966, Dr. Henry K. Beecher, an anesthesiologist, wrote an article (Beecher HK. “Ethics and Clinical Research” NEJM June 16, 1966) describing 22 examples of research studies with controversial ethics that had been conducted by reputable researchers and published in major journals. Beecher’s article played an important role in heightening the awareness of researchers, the public, and the press to the problem of unethical human subjects research.

One of the seminal events in the development of the current regulatory environment was the Public Health Service (PHS) Syphilis Study (1932 – 1972), the so-called “Tuskeegee Syphilis Study”. Initiated and funded by the PHS, this study was designed to document the natural history of syphilis in African-American men. Hundreds of poor, African-American men with syphilis were enrolled into the study. The men were recruited without informed consent and were deliberately misinformed about the need for some of the procedures. This longitudinal study lasted over 40 years until newspaper reports forced the US government to terminate the study. For more information follow the link to the PHS Syphilis Study.

1.2 Events in Social & Behavioral Research

Events contributing to the development of the current regulatory system were not limited to biomedical research; during the same period there were several similar events in the social and behavioral sciences: The Wichita Jury Case (1953) where researchers tape recorded jurors’ deliberations in six cases to measure influence of attorney comments on decision making. The research was conducted with knowledge of the judge and attorneys, but not jurors. The Milgram “Obedience to Authority”(1963) studies which were conducted to determine how far subjects would go in administering seemingly severe electric “shock” as directed/instructed by an authority figure (to continue when the experimenter) to another subject (a confederate) even when the latter subject appeared to be in extreme pain but continued to fail test questions. Humphreys “Tearoom Trade” study (1970), which involved the observation of men engaged in sex acts in restrooms, secretly following them to their cars, transcribing license plate numbers, tracking them through DMV records to their homes and interviewing them about personal issues. The Zimbardo “Simulated Prison” (1973) research, which involved assigning roles to male student volunteers as “prisoners” and “guards”. The research became so intense as physical and psychological abuse of “prisoners” by “guards” escalated, that the researcher stopped the experiment/simulation after six days. See Dr. Zimbardo’s web site for more details on this study.

UPDATE (7/8/2011): I heard back from CITI.