送交者: mangolasi 于 2006-4-27, 14:47:50:
回答: Re:人也不是理性的动物 由 mangolasi 于 2006-4-27, 10:51:27:
If the people are making decision AT THE BEGINING whether to attend a game with 20 trials, using Law of Large number, you can approximately think the expected gain is $1.5 and it is quite certain (small variation). In that sense, not attending it is irrational.
However, after reading the article carefully, I found the guys are asked to decide whether to attend EACH TRIAL, and xj's explanation, though unclear in concepts, is more or less get the point. When you have $5, losing $1 is a big deal and while you have $30, losing $1 is worthy for the potential of getting $2.5. So the "losers" will be more reluctant to gamble, rationally. Though I don't know what happen in the damaged people--and there is certainly something about behavior going on there, we can not disintangle the two effect: "rational" risk-aversion and irrational choice (pressure of losing) quantitatively. The latter is not identified. The whole experiemnt need to be redesigned if you want to give out more than merely QUALITATIVE IMPRESSION.