5 ms·

I think the most important thing to consider here is that these logic questions they use to test "rationality" are awkward word puzzles, and the test reveals mo

I think the most important thing to consider here is that these logic questions they use to test "rationality" are awkward word puzzles, and the test reveals more about how well people can work out what's being asked than their underlying rationality. The fact that most people would jump onto the "bank teller and feminist" option in the conjunction fallacy test is because the question as posed contains mainly feminist-related ideas, and so the brain's natural tendency is to jump to the answer containing the word feminism. That's less about rationality and more about how attuned are you to the likelihood that the questions you're being asked are worded in a tricky way.

gwern 12y ago
> I think the most important thing to consider here is that these logic questions they use to test "rationality" are awkward word puzzles They aren't. Have you looked at the CRT questions? There's nothing tricky about them, the challenge is simply actually checking your answer. And if it's 'just' a word puzzle, why is the correlation with IQ so low? And why does this sort of rationality inversely correlate with screwing your life up in various ways (see the research discussed in my comment in OP)? That's an odd thing for an 'awkward word puzzle' to do. > The fact that most people would jump onto the "bank teller and feminist" option in the conjunction fallacy test is because the question as posed contains mainly feminist-related ideas, and so the brain's natural tendency is to jump to the answer containing the word feminism. You know that was the very first criticism leveled against Kahneman & Tversky's conjunction fallacy results and they long ago showed that other wordings gave the fallacy too?