4 ms·

There is some truth here, especially in regards to more intelligent people taking a mainstream view that lacks the illusion of nuanced thought. That said, I fi

There is some truth here, especially in regards to more intelligent people taking a mainstream view that lacks the illusion of nuanced thought. That said, I find it hard to read an article on "intellectual hipsterism" that claims the average IQ of their readers is 145. The page the author links to in support of that claim mentions two giant caveats: First, IQ is a notoriously difficult to measure metric of questionable validity, and second, the scores are self-reported. The reported values range from 120-180. Essentially, the survey suggests that the top half of their readership is in the top 1% of the national population. Even their bottom half is in the top 10%, by their statistics. This suggests a fatal flaw in the survey, and to mention the resulting statistic in such an offhand manner and with no warning to the reader of the problems with it suggests at best laziness, and at worst intellectual dishonesty on the part of the author. I find it hard, therefore, to take their argument at face value. That's not to say it's invalid, merely that the author's conduct does not support its validity.

fjh 16y ago
To be fair, there is a footnote pointing out how dubious that claim is.
- jbooth 16y ago
  Yeah, I read the sentence as tongue-in-cheek humor.
  - redthrowaway 16y ago
    I think that's a fair analysis. It wasn't how I took it, but I could see that being the author's intent.
nkurz 16y ago
I would take it at face value. First, the author is taking for granted that his readers are intelligent enough to realize that self-reported IQ scores are inaccurate, that IQ as measured is not equivalent to any real world intelligence, and that the accuracy of IQ tests is limited once one gets out to the extremes. He links directly to the survey --- what more can he do? Second, I'd guess that the numbers are about right. In the same way that one would be safe assuming that the posters to StackOverflow are probably mostly in the top 1% of programming ability out the entire population of the world, I think it's likely a site dedicated to debating 'meta-contrarianism' might select quite precisely for people who can score well on standardized tests. I'm not a regular reader, but I know enough about the site to suggest a solution: come up with a means of testing your hypothesis, and put up some money to back it. The more direct the better: bet them $1000 that none of the regular posters can show a bona-fide IQ test result higher than 160 (or whatever cutoff you think fairly accounts for the limited accuracy of the test at the extreme). And then report back to us if it was money well spent!
- redthrowaway 16y ago
  I'd rather not, thanks ;) I'm sure if I challenged someone to come up with a result of 170, there'd be at least one person out there who could do it. My money is far better spent on tuition and beer. Now, if I was a psych major, I might be able to get approval for a study comparing the congruency of self-reported IQ scores with measured, neutral ones in self-described intelligent communities. I'd apply for funding to set up testing for those who took the initial survey, and see what the results are.
  - nkurz 16y ago
    That's a good plan too. They'd probably go for it. In the commentary on the survey, Eliezer (founder of the site, also present here) doubts the estimate too. It's worth noting that one of the central tenets of LessWrong is that it can be just as bad to underestimate one's knowledge as to overestimate it. As a result, some things which look like boasting are really just attempts at accuracy, carried through to a level not normally seen in polite discourse. Other things, of course, are just boasting.

fjh 16y ago

To be fair, there is a footnote pointing out how dubious that claim is.

jbooth 16y ago
Yeah, I read the sentence as tongue-in-cheek humor.
- redthrowaway 16y ago
  I think that's a fair analysis. It wasn't how I took it, but I could see that being the author's intent.

nkurz 16y ago

I would take it at face value. First, the author is taking for granted that his readers are intelligent enough to realize that self-reported IQ scores are inaccurate, that IQ as measured is not equivalent to any real world intelligence, and that the accuracy of IQ tests is limited once one gets out to the extremes. He links directly to the survey --- what more can he do? Second, I'd guess that the numbers are about right. In the same way that one would be safe assuming that the posters to StackOverflow are probably mostly in the top 1% of programming ability out the entire population of the world, I think it's likely a site dedicated to debating 'meta-contrarianism' might select quite precisely for people who can score well on standardized tests. I'm not a regular reader, but I know enough about the site to suggest a solution: come up with a means of testing your hypothesis, and put up some money to back it. The more direct the better: bet them $1000 that none of the regular posters can show a bona-fide IQ test result higher than 160 (or whatever cutoff you think fairly accounts for the limited accuracy of the test at the extreme). And then report back to us if it was money well spent!

redthrowaway 16y ago
I'd rather not, thanks ;) I'm sure if I challenged someone to come up with a result of 170, there'd be at least one person out there who could do it. My money is far better spent on tuition and beer. Now, if I was a psych major, I might be able to get approval for a study comparing the congruency of self-reported IQ scores with measured, neutral ones in self-described intelligent communities. I'd apply for funding to set up testing for those who took the initial survey, and see what the results are.
- nkurz 16y ago
  That's a good plan too. They'd probably go for it. In the commentary on the survey, Eliezer (founder of the site, also present here) doubts the estimate too. It's worth noting that one of the central tenets of LessWrong is that it can be just as bad to underestimate one's knowledge as to overestimate it. As a result, some things which look like boasting are really just attempts at accuracy, carried through to a level not normally seen in polite discourse. Other things, of course, are just boasting.