Dec 19 2008

Milgram’s Famous Studies Finally Replicated

Published by under Uncategorized
Comments: 22

In the 1960s and 70s Stanley Milgram performed a series of now famous obedience studies. The most famous of which, experiment 5, involved a subject being asked to deliver shocks to a confederate (someone who was in on the experiment) in what they were told was a learning experiment. Surprisingly, 65% of subjects continued to deliver shocks all the way to the end, even over the increasingly vehement protests and even medical complaints of the confederate. Prior to Milgram most people, and even psychiatrists, did not predict such an outcome. Almost no one thinks they personaly would behave like the subjects in experiment 5, but apparently 65% of us are wrong. These experiments stand as a dramatic demonstration of the power of authority and other situational factors in human behavior.

Milgram’s experiments, however, have been controversial. There is always, of course, controversy over exactly how to interpret social psychology experiments. Human behavior is extremely complex, and so there are always numerous variables to consider when interpreting such studies. But even more controversial than the interpretation are the ethical considerations raised by the study. Specifically, the subjects were exposed to significant short term stress during the study. Ethical guidelines since Milgram have made it impossible for anyone to replicate his studies over the last 30 years, which means there is no way to end the controversy over interpretation.

However, researcher Jerry M. Burger has now published a fairly decent replication of the famous experiment 5, staying within current ethical guidelines. The major difference in his study design is this: In the Milgram studies, which involved increasing the voltage each time the confederate gave a wrong answer or refused to answer, most of the participants paused at the 150 volt level, which was when the confederate first gave an audible cry of pain. However, 79% of those who gave the fake 150 volt shock went all the way to the end of the experiment, giving the maximum 450 volt shock three times. Burger therefore reasoned that he could end his replication at the 150 volt shock level and that would be sufficient, and 4 out of 5 subjects who gave that shock went all the way, so their behavior could reasonably be predicted. He also put other ethical controls in place, such as screening subjects more thoroughly and rapidly briefing them on the true nature of the experiment when it was over.

Burger found that 70% of subjects delivered the 150 volt shock, compared to 82.5% for Milgram. This difference was not statistically significant. This would appear to replicate Milgram’s findings. But of course, Burger’s protocol, which is being called “obedience lite” is not an exact replication, so I’m sure the debated will ensue.

There was one difference in Burger’s results. Milgram found that if he modeled disobedience by exposing the subjects to another “subject” (actually a confederate) who refused to proceed, then the subject would be more likely to refuse. Burger did not find this effect, although he used different defiance modeling and so direct comparisons are difficult.

The most widely known interpretation of the Milgram experiments is that it demonstrates the effects of obedience to authority. Subjects proceeded to give more and more intense shocks at the direct prodding of an authority figure – the scientist conducting the experiment. However, there are potentially other factors as well. The gradual increase in demands makes it easier for subjects to proceed. Shocks were allegedly given in 15 volt increments. Giving smaller shocks at the beginning of the experiment may make them more accepting of the overall procedure. Also, it may seem inconsistent to refuse to give a 150 volt shock when you just gave a 135 volt shock.

Also, having limited sources of information in a novel or uncertain situation is also a factor. Other psychological experiments have shown that being uncertain in a situation can be morally paralyzing. Subjects in an unusual situation with an authority figure telling them what to do can be a powerful combination.  And finally, in the experimental situation responsibility was diffuse and not assigned.

Although Burger’s experiment was not an exact replication, he argues that it is a reasonable model and allows exploration of the same psychological phenomena as Milgram studied, and this is reasonable. His findings add to the bottom line interpretation of such experiments – situational factors can have a profound effect on human behavior, and we tend to grossly underestimate the effect of situational factors, especially on ourselves. Understanding this can be a very useful insight. It may help explain events such as the abuses that took place at Abu Graib. Such insight may even help us avoid similar situations in the future.

22 responses so far