The trusted source for
healthcare information and
Famous obedience study recreated – at least in part
Changes to reduce subjects stress win IRB approval
The 1960s obedience experiments of Stanley Milgram have become a hallmark of social behavioral research, as well as a cautionary tale for those involved in human subjects protection.
Milgram showed that ordinary people could be induced to deliver what they believed were painful, potentially severe electric shocks to others at the direction of an authority figure. While the results were dramatic and illuminating, they also raised questions about the treatment of the subjects, who were subjected to the stress of hearing what they believed to be screams of pain and pleading.
Among researchers, Milgram often has been seen as an experiment that couldn't be done today, in part because no IRB would allow it.
Jerry Burger, PhD, a social psychology researcher at Santa Clara University in California, was among the doubters. But when ABC News approached him in the aftermath of the Abu Ghraib prison scandal to see if a replication of the Milgram study might be designed to help explain the behavior of prison guards, he decided to try.
The result was a scaled-back version that eliminated many of the objectionable details of the Milgram studies, while arriving at nearly the same conclusion.
"My findings indicate that the same situational factors that affected obedience in
Milgram's participants still operate today," Burger wrote in a January 2009 article in the journal American Psychologist.
He says he knew the study would be a tough sell to his IRB, which normally doesn't tackle subjects with such potential for controversy.
"The easy thing for them to do would be to say no," says Burger, who himself had previously served on the board. "If I was on an IRB and I got this proposal, I think my first inclination would be to say 'This is impossible.' Fortunately they didn't do that, and they gave it careful consideration."
In designing the study, Burger made a half-dozen important changes to the original Milgram experiment (see accompanying story), all designed to lessen stress on participants. Most notably, he cut off the experiment before the "shocks" they delivered (actors actually feigned responses to non-existent shocks) reached a level that participants might see as severe or distressing.
Before he even attempted to submit a protocol to the IRB, Burger sought out the advice of others in his field.
"I consulted with people who were experts on Milgram and experts on the ethics involved," he says. "I wanted to make sure I wasn't off base here – if they had all told me, 'No, no, that's crazy, you can't do that,' then I would have stopped at that point. But their judgment was that would probably be an ethical way to get some relevant information. That was encouraging."
One of those he sought out was Steven Breckler, PhD, a social psychologist who is executive director of the Science Directorate of the American Psychological Association in Washington, D.C. Breckler has written about the ethics of the Milgram experiments.
"Obviously the APA doesn't approve or disapprove of studies, that's not their role," Burger says. "But he wrote a nice letter for me that I was able to pass along to the committee with his permission, saying he had looked this over and thinks it is within the guidelines."
Burger says he thought Breckler's support was reassuring to his IRB.
"If I was on this committee and I wasn't sure, seeing that this person who's obviously in a very responsible and knowledgeable position has made this assessment – I think that made it easier for them."
Burger also offered to provide the IRB with a list of experts whom they could consult about the study.
"I tried to anticipate what all their objections might be," he says. "I gave them a very lengthy description of the procedure, and they met many hours discussing all the fine points. In the end, they told me they had decided right away it was probably okay to do the study."
Television broadcast issues
The IRB did ask for two changes. The first related to the potential for video from the experiment being broadcast on television. While all subjects had to give permission for the use of their images after they were told the true purpose of the study, the IRB wanted more assurance that a participant wouldn't give permission and later change his mind. As a result, participants were contacted by phone and by mail afterward, to make sure they did not want to withdraw permission.
"In fact, nobody did," he says. "Something like 90% of the people eagerly signed saying yes, they were willing to be on television and even though they were given multiple opportunities to take back that permission, they didn't."
The second point raised by the IRB was the use of clinical psychologists to screen potential participants to identify anyone who might be unduly stressed by the study. The IRB asked that the psychologists chosen be people that Burger didn't know personally in order to avoid potential conflicts of interest.
Breckler says he wasn't surprised that the IRB ended up approving Burger's replication.
"I think some IRBs probably would not approve it, but I do think his procedural solutions minimized the risk to participants," he says.
Breckler says Burger's experiment may be a good case lesson in how an IRB can balance risks against potential benefit in social behavioral research. He notes that Milgram's original experiment was an attempt to understand the behavior of Nazis during the Holocaust, and that Burger's replication came in response to prisoner abuses at Abu Ghraib.
"I think you could see this study as an example where the potential benefits of understanding this behavior could outweigh the risks by a huge amount," particularly when there's been an effort to minimize those risks as much as possible, Breckler says.
He says the American Psychological Association encourages researchers to collaborate with IRBs in this way when they're setting up studies.
"You like to hear stories like this – it's refreshing to see the ability of IRBs and researchers to engage in a partnership," he says.