Survey taken, answered all questions I could. This excluded the IQ question set. I’ve never taken an IQ test. I’ve never been offered an IQ test, nor considered taking one. Is that strange? The survey seemed pretty confident that I’d have measured my IQ.
A previous incarnation of the test just asked what your IQ was. We got both people who had taken official tests responding, and people who were just estimating their IQ. The second group is really noisy, and made it difficult to meaningfully talk about the IQ of LWers.
I suggested the current question as a way to get high-quality information out of survey-takers, but I also wanted a question where people estimated their IQ (maybe as two questions, for the lower and upper bound of a 50% CI) so that we could still get the low-quality information.
Respondents with high IQ seem more likely to have taken official tests, though; doesn’t this overestimate LW’s mean?
Any self-report will overestimate LW’s mean, even if there is no disproportionality among test-takers. I’ve taken this into account with various assumed population means in the analysis of previous surveys, but there’s fudging involved (if the average IQ of responders is 130, is it really sensible to expect non-responders have an average IQ of 100?).
However in the US the usual standardized tests (SAT, GRE, GMAT, LSAT, MCAT) are highly correlated with IQ and going by percentiles you can get a reasonable IQ estimate easily enough.
My point is that the renorming in the 1990s (if I remember correctly) chopped off the right tail of the SAT distribution. It used to be that about 1 in 4000 people got SATs of 1600, and so that implied a commensurately high IQ, but now about 1 in 300 do (only looking at M+CR), so the highest IQ level that the SAT is sensitive to has dropped significantly.
If I remember correctly from the last year’s survey, the mean SAT score of LWers who reported it implied that the mean LWer was about 98th percentile, which seemed about right to me (and suggests that the SAT is a decent tool at discriminating between most LWers).
I haven’t taken any official IQ test nor have I taken any standardized tests. The only sort of official intelligence test I took was the ASVAB, though I forgot what my score was. I did score high enough to take the DLAB though (I was originally tasked to be a Turkish linguist in the Air Force).
Survey taken, answered all questions I could. This excluded the IQ question set. I’ve never taken an IQ test. I’ve never been offered an IQ test, nor considered taking one. Is that strange? The survey seemed pretty confident that I’d have measured my IQ.
A previous incarnation of the test just asked what your IQ was. We got both people who had taken official tests responding, and people who were just estimating their IQ. The second group is really noisy, and made it difficult to meaningfully talk about the IQ of LWers.
I suggested the current question as a way to get high-quality information out of survey-takers, but I also wanted a question where people estimated their IQ (maybe as two questions, for the lower and upper bound of a 50% CI) so that we could still get the low-quality information.
I consider giqtest.com also professional/scientific, despite being taken online.
(I understand the general aversion towards online tests, and don’t mind the current wording.)
Respondents with high IQ seem more likely to have taken official tests, though; doesn’t this overestimate LW’s mean?
Any self-report will overestimate LW’s mean, even if there is no disproportionality among test-takers. I’ve taken this into account with various assumed population means in the analysis of previous surveys, but there’s fudging involved (if the average IQ of responders is 130, is it really sensible to expect non-responders have an average IQ of 100?).
I’ve never taken an IQ test either.
However in the US the usual standardized tests (SAT, GRE, GMAT, LSAT, MCAT) are highly correlated with IQ and going by percentiles you can get a reasonable IQ estimate easily enough.
This is no longer true for high IQs, and most of the conversion tables are only for the old SAT. A 1600 just ain’t what it used to be.
Measuring high IQs is difficult in general, but a rough estimate on the basis of, say, SAT scores is better than no data at all.
My point is that the renorming in the 1990s (if I remember correctly) chopped off the right tail of the SAT distribution. It used to be that about 1 in 4000 people got SATs of 1600, and so that implied a commensurately high IQ, but now about 1 in 300 do (only looking at M+CR), so the highest IQ level that the SAT is sensitive to has dropped significantly.
If I remember correctly from the last year’s survey, the mean SAT score of LWers who reported it implied that the mean LWer was about 98th percentile, which seemed about right to me (and suggests that the SAT is a decent tool at discriminating between most LWers).
I haven’t taken any official IQ test nor have I taken any standardized tests. The only sort of official intelligence test I took was the ASVAB, though I forgot what my score was. I did score high enough to take the DLAB though (I was originally tasked to be a Turkish linguist in the Air Force).