Apocryphal psych experiments

There are two unrelated psychology experiments that I heard about long ago that are funny and unexpected in what they reveal about human nature, at least if the distant accounts are anywhere close to being correct. But of course this kind of tale always gets better in the telling, and both of these tales had been told many times before they got to me. So I’ll tell them again (and hopefully they’ll be even better this time), but I’m also interested in being pointed to the original studies if anyone can recognize them from these stories. The first is about memory — the story is that the experimenters were interested in to what extent people could accurately assess whether they had good memories or not. They were surprised to find an _inverse_ correlation between memory skills and self-assessed skills — that is, people who demonstrated good memories in experimental tests described themselves as having bad memories, and vice versa. The eventual just-so explanation was this: people with good memories can remember many instances of having forgotten things; people with bad memories can’t remember _ever_ having forgotten anything …

The second was about depression — the story is that the experimenters wanted to study the extent to which depressed people have an inaccurate and inappropriate feeling that they are not in control of events. So they devised an experimental game where the outcome was partially random and partially under the control of the subject, and asked the subjects to play, and then to assess afterwards how much control they had actually enjoyed. Two pools of subjects played – a depressed group and a control (not-depressed) group. The surprising result was that the depressed subjects were quite accurate in their estimates about how much control they had; the not-depressed subjects overestimated their control wildly. Inferences about the true meaning of “mental health” and the human condition are left to the reader…

If either of these rings a bell, please leave a comment, or send me mail. (My email address is easy to discover by going to http://timconverse.com/, although you might have to hit refresh a couple of times and then make an inference or two (which should be easy if you’re human).)