Friday, December 19, 2008

Musings on Milgram

I was just reading about the replication of the Milgram experiments, which show that people are surprisingly obedient to authority. The majority of subjects will deliver painful shocks to others when a researcher instructs them to do so. This of course has broad implications in any situation involving authority figures, particularly in regards to all the torture controversy of late - who is really responsible?

A couple of thoughts came to mind.

I would love to see a follow-up, similar to my suggestion for acupuncture research, to study whether the obedience to authority is any weaker for subjects who know about the Milgram experiments. I'm not so naive to think that I'm fundamentally more benevolent or resistant to authority than average, but I would like to think that in the same situation, knowing about the Milgram experiments would help me overcome the effect. My guess is that it would, in the same way that knowing about common logical fallacies, susceptibility to illusions and false pattern recognition, etc. help me avoid falling for pseudoscience and hoaxes.

This got me thinking about psychological studies in general. In the interest of proper blinding, test subjects are almost always mislead regarding the true nature of the experiment (and filled in afterwards, of course). Milgram told test subjects they were testing shock aversion in a learning experiment.

If you know this is standard practice, and you suspect that you're actually participating in a different test than the one overtly described, won't that alter your behaviour? You'd be on your guard, suspicious of everything. Like a suspect being interrogated, you won't be fooled by the big mirror on the wall. To what extent does that invalidate the experiment? How do you filter that out? As scientific literacy increases in society (fingers crossed) will this become a major problem for psychological and behavioural research? How might they overcome or work around the problem?


Mozglubov said...

I took a psych course a year and a half ago, and part of our mark was dependent on being the subject of at least two studies. I was really on the lookout for exciting twists and turns and profound psychological conspiracies. Unfortunately, nothing of that sort happened... the most exciting thing was I got to do a brief exercise in 3D block rotation in which I discovered I am apparently rather good at mental block rotation. It's always nice to find out you are good at something.

I do remember, though, thinking that all my alertness for tricks and self-reflection caused by the very fact that I was participating in a psychology experiment were mitigating factors for whatever results came out. I'm not really sure how one would get around that, though. Maybe we need to adjust our ethical outlooks on surveillance...

Jolly Bloger said...

Yeah, I guess even without suspicion of subterfuge, the very fact that you're being tested will alter behaviour. There must be a standard treatment of this, they must know how that affects different kinds of tests.