Just read up on Stanford Prison Experiment. What am I missing?

[wickerwoman]

OK, I haven't studied psychology but I have done research in other fields. It seems to me that this experiment has massive problems with its methadology.

I've seen a number of people point to this study to explain the behavior of the guards at Abu Ghraib and to suggest that under the right circumstance we could all become monsters.

Reading the study, I don't see it.

Start with the tiny, tiny sample size. The researchers used 18 people. 9 guards and 9 prisoners. The subjects were not chosen randomly, they were affluent middle-class white Ivy-League college men. The study describes them as "healthy" and "normal" as a result of psychological screening. Assuming these are 1971 code words for "straight" I will assume that they were also nominally heterosexual.

I'm not going to get into any essentialist garbage about men being naturally more violent than women, but I would like to point out the probability that people with this background in 1971 were almost certainly sheltered from the kind of violence and discrimination that people of color, women and gays faced and continue to face- people who have never been victimized and as such are less likely (at 18-21) to have developed a keen sense of empathy for victims of harrassment and discrimination.

Over the course of the study, the prisoners began to lose their sense of self and conform to the expectation of the guards. And they were abused by some guards who thought they were free from observation.

But the result of the study found that three of the guards ignored the problems, three of the guards empathized with the prisoners and made nice with them, and three of the guards went power mad and instigated the majority of the violence and harrassment. So in 33% of the (9) cases absolute power corrupted absolutely. This is actually far less than I would have expected.

I'm sorry, but this just isn't sufficient evidence that a majority of people put in the same situation as the guards at Abu Ghraib would respond in the same way. And what is the implication even if it was true? Are they then absolved from responsibility for their actions?

I would love to hear from people who have studied psychology or this case in more detail in case there is something I misunderstood or didn't see. What is the current take on the findings of this study?

[Wickerman]

1. You've got the fundamentals right

The sample is too flawed to draw true conclusions. More often the experiment is used to show how not to conduct an experiment - why you need to stick to controlled environments, human subjects review, etc.

I think a better study to reference would be the discrimination studies conducted in small town Iowa decades ago:
http://www.llnl.gov/llnl/06news/Employee/articles/2002/04-19-02-elliott.html

A similar dynamic is at play - someone is told they are better than the other group, both groups react in kind. That study has replication behind it; the Standford prison experiments were pretty nasty and cause enough not to replicate them.

Lessons could be learned from the SPE, but not the ones the media wants to cram down our throat. A colleague and I argued this at work just yesterday - the premise being that it could be the SPE coming to life - me saying hte SPE was too flawed to be certain. Actually, its probably true, but you can never know and it should be treated as a never know if you want to view it scientifically.

[pop goes the weasel]

2. the number goes up

when people believe they are acting under orders.

See this summary or this one of Obedience to Authority for more information. Better yet, read Stanley Milgram's Obedience to Authority.