Showing posts with label milgram. Show all posts
Showing posts with label milgram. Show all posts

Friday, January 29, 2010

MILGRAM'S FATALLY FLAWED EXPERIMENT


I'm afraid I have only a limited respect for psychology. The field attracts too many quacks and sloppy thinkers. One of the most famous sloppy thinkers was Stanley Milgram whose famous experiment in the 6os was said to have proved that a large number of people are capable of cruelty when authorities sanction it. That's certainly possible, I just don't think Milgram proved it, or even came close to proving it.



Could Milgram's experiment have been as obviously flawed as this BBC re-do of the experiment? Here the victim's screaming voice is obviously fake. It sounds speeded up and is obviously pre-recorded to boot. Most of the students administering the charge must sense, at least unconsciously, that something's up, and the situation isn't to be taken seriously.

The controls in front of the scientist (who's dressed like a doctor, and acts like one) are never explained, and are plausibly misunderstood by the student to be a means of modifying the charge or keeping track of the subject's vital signs. It shouldn't be surprising that that most of the students weren't terribly worried about the health of someone who's under close medical supervision.



Add this to the fact that in the lawyered-up age we live in everybody knows that a university would never allow a casual test subject to be put in serious jeopardy. Further add that the researcher continuously assures the student that no lasting injury will be done to the subject, who can quit at any time. The student administering the charge has to conclude that the screaming man is probably not in real jeopardy. The fact that 30% of the students still refused to hurt an obvious or near-obvious fake could actually be taken as a ringing endorsement for the goodness in man.

A good experiment is one that definitely confirms or excludes an explanation and this experiment does neither. I'm amazed that Milgram didn't realize that, and even more amazed that subsequent academics failed to see it.



Milgram was an odd person. He's also famous for the "Lost Letter Experiment" in which he demonstrated that found mail addressed to unpopular people like famous nazis were less likely to be re-mailed than letters addressed to normal people. I hope my taxes didn't have to pay for that.

Psychology does have a place, but it seems like its misused as often as not. It's very trendy, very faddish. Here's (above) a cover of Psychology Today from 1974. It informs us that psychosurgery (remember the fake Philipino guys that removed chicken entrails from people's stomachs without breaking the skin?) is on the level, and that Yuri Geller (a magician who claimed to be a psychic, exposed by Randi ) is on the up and up. Other issues touted open marriage (how many of these ended in divorce?), kibbutz-style raising of children (since dropped by Israel), and a bunch of other later discredited ideas. This was an influential magazine in its day. What does that tell you about standards in the field of psychology?

As a footnote, here's a link to 5 experiments which the author says prove that humanity is doomed. Milgram's is number one. I liked the article myself, but then again, I don't take things like this too seriously. Most isolate the unsocial or uncaring thing the subjects did when tested, and ignore the social and caring things the subjects did during the rest of the day. A fun read, but not very good science.