The trusted source for
healthcare information and
Comparing two studies: A painful balancing act
Participants still believed they were delivering shocks
In trying to replicate the Milgram obedience experiments, social psychologist Jerry Burger had to balance the goal of getting useful results that were roughly comparable with Milgram's, while providing more protection for participants from severe stress.
Stanley Milgram actually conducted a whole series of obedience studies, but most were variations on his first experiment, conducted in 1961 at Yale University. /Prompted by Nazi war crimes trials, Milgram wanted to explore how people could follow orders to commit such atrocities.
He devised an experiment in which a participant was assigned the role of "teacher," and told that the study looked at the effects of punishment on learning. The participant watched as a "learner" (actually a confederate of the researcher) was strapped into a chair in an adjacent room and electrodes were attached to his body.
The participant was told to administer a test to the learner, and to deliver incrementally more powerful electric shocks when he answered incorrectly.
As the dial was turned to higher settings, the participant could hear the learner's cries of protest and demands to be let out. Meanwhile, an authority figure in the room would tell the participant to continue. The experiment would continue until the participant repeatedly resisted instructions to keep delivering shocks. At the highest settings, the cries of protest would end, suggesting that the learner no longer was capable of responding.
The shocks, of course, were not real.
Milgram found that 65% of participants were willing to continue delivering what they believed to be severe electric shocks, all the way to the end of the machine's range, at 450 volts.
"I've seen the film that Milgram produced, and common sense tells you that this is way too stressful to put anybody through," says Berger, PhD, a social psychology researcher at Santa Clara University in California. "If you've ever seen the images and heard the screams coming through the wall – a man screaming 'Let me out! I can't stand the pain!'
"It is unbelievable. I knew it would never be acceptable – it wouldn't be acceptable to an IRB, but it also wouldn't be acceptable to me."
Screening subjects, dialing down shocks
Burger took a number of steps to protect subjects:
A 'personal standard'
Burger says that as he devised his study, he tried to keep in mind how he would feel if his own grown son were participating.
"Would that be fine with me? And if a researcher asks that question and says, no, I don't think I'd like that, then I think we have no business putting anybody else in that study. I think that's a pretty good barometer to use. Even if the IRB approves it, there's still the personal standard that everybody needs to live up to."
He says he's been contacted by other investigators about the possibility of replicating his study and he cautions them to consider two points carefully. The first is the amount of stress participants are subjected to, and the possibility that they could have a negative experience.
The other point, he says, is trickier – the conflict between a participant's right to leave the study at any time, and the necessary actions taken in the obedience study.
"Ordinarily, if somebody says, 'I don't want to do this anymore,' I always tell my experimenters that the study is over at that point. You don't coerce people into staying if they say they don't want to be there," Burger says. "But in this Milgram replication (as in the original), we didn't immediately let them go; we made them say repeatedly that they wanted to stop. And even though this may only last for a few seconds, it gets back to the question of whether we're putting people through something that you shouldn't.
"It's a tough question, and the lingering question for me that I still wrestle with. In this case, given how it worked out, nobody was upset that they weren't released as soon as they said they were uncomfortable. But that's something I think researchers need to think about if they're going to do something like this."