A new variation on the Asch conformity experiment was recently published. The experiment was performed in Japan and used polarizing glasses to show different lines to different people in the same room, so that the subjects had to disagree with others they actually knew, and who genuinely believed that they were answering correctly. The study found that women conformed by giving a wrong answer about a third of the time, but men did not.
Learned about this via Ben Goldacre's blog.
Interesting. That's a clever solution to the confederate problem.
Personally, I know I'm never quite prepared to discard the possibility that I am completely out of my mind and/or physics might behave in a drastically unexpected way any moment. Obviously I don't assign this a high probability, but higher than average, I think. So in situations where everyone else seems to agree about something that seems wrong to me--especially something arbitrary/trivial, as it would almost necessarily be in an ethical test situation--it wouldn't take much to get me to assume that I'm simply the one who's wrong.