3 ms·

Nobel winner James Heckman did a lot of work on this [1]. The takeaway is that IQ doesn't matter as much as other attributes (character, learned skills, etc.)

Nobel winner James Heckman did a lot of work on this [1]. The takeaway is that IQ doesn't matter as much as other attributes (character, learned skills, etc.) in life outcomes. HN style crows way overvalue IQ style skills when intelligence has multiple dimensions, many of which matter more than IQ. IQ is effectively just a measure of test-taking skills. That's useful in some areas, but you could see someone like Gengis Khan or Nikola Tesla be incredibly intelligent and a poor test taker. [1] https://heckmanequation.org/resource/i-q-isnt-everything/ https://heckmanequation.org/resource/i-q-isnt-everything/

tgv 4y ago
IQ is a fair predictor of achievement in life, though. But there are other predictors too, such as parents' socio-economic status, or geographical location. We wouldn't claim these things measure intelligence, though. IQ is a proxy for intelligence, but one with great shortcomings. Unfortunately, there's no agreed upon definition of intelligence, let alone an operalization, and a way of measuring it. So IQ keeps being used.
- mjburgess 4y ago
  > Unfortunately, there's no agreed upon definition of intelligence, let alone an operalization, and a way of measuring it. So IQ keeps being used. Uhuh.... so this line it's crazy, right? It's really strange you concede that we have no theory and hence no way of verifying whether an IQ score is a valid and reliable measure of the thing we have no theory of... and yet, you say it is a "proxy for intelligence" The entire edifice of research which makes these claims is a product of a pseudoscientific use of statistics by amateur statisticians. The vast majority of the results you think exist in the literature are frauds, and often egregious ones at that. A human being is one of the most complex systems known to exist in nature, and the frauds in psychometrics are very far away from being up to the task of studying them. The greatest minds in applied sciences, matehmatics and statistics would largely throw up their hands at the job of measuring "intelligence" --- that some fools havent, is no good thing.
  - tgv 4y ago
    We all "sort of" recognize intelligence, but it stops there. > The entire edifice of research which makes these claims is a product of a pseudoscientific use of statistics by amateur statisticians. That's bollocks. It's just a topic that's too complex. Unfortunately, everybody likes it and draws conclusions from it. Even the HR department. If you would say we should ignore IQ measurements in society, I would agree. But it's not pseudo-scientific use of statistics. Nor all done by amateurs. It might even be more scientific than string theory or dark matter.
    - mjburgess 4y ago
      Alas, it is. It's severely pseudoscientific. None have PhDs in statistics. They're all people who got into psychology and were given GUIs by the maths department. Their papers are a symptom of what happens when you press the "ANONVA" button and print the results. The entire field is an egregious testament to pseudo-intellectual pseudoscience. It's shocking. There's nothing there. It's unfathomably bad.
      - tgv 4y ago
        The statisticians I knew were only too happy to go along with some ANOVA. They're usually consulted at the start of the study, but rarely co-author. A lot of psychology is bad research. There's a reason there's a reproduction crisis. Statisticians are to blame for that, too. They've been teaching and using NHST, even though all arguments point away from it. But statisticians can't fix a bad design, because they rarely know the factors involved, and they can't fix errors in the methods, because they've got no idea about the subject matter. The statistician in your argument comes across as some kind of totem. > There's nothing there. There's something there, but since we don't know what, it's best to take it with a grain of salt. However, if even the bloody HR department looks at IQ scores, it gets a lot of attention. And one explanation is that it predicts better than nothing. If you have to choose between two candidates, and one has an IQ of 80 and the other of 120, you know who (on average) who fits best.
        4y ago
ilyt 4y ago
"test-taking" is too generic. Puzzle-solving would be more accurate. IQ doesn't do much for test that just require to remember a lot of stuff.
maximus-decimus 4y ago
But IQ wasn't invented to predict "life outcome", it was invented to find kid's academic potential. I'm still waiting for somebody to show me a prolific Math Phd with below average IQ. The only counter example I've heard people use it Feynman whose IQ was "only" 125 putting him smarter than "only" 95% of the population. It rubs me the wrong way when Hawkins says stuff like "IQ is for losers" but then boasts how his physics undergrad degree was so easy for him he was just bored out of his mind (it sure wasn't easy for me). It's the intelligence equivalent of a billionaire saying you just need to work hard to get there, as if that was ever gonna make a difference if you're an Amazon factory worker.
- mjburgess 4y ago
  It was invented as a clinical test of retardation, and the only correlation with IQ and objective independent measures of performance is in the region of very low scores. Note, that when carear-IQ correlations are brought up it is universally in those populations subject to an IQ test filter who are then shown to correlate with IQ scores (ie., those who take, eg., MCAT, LSAT, ...) --- which, if it isnt clear, is circular. Average and above IQ shows no correlation whatsoever with, eg., wealth.
- P5fRxh5kUvp2th 4y ago
  I've never taken an IQ test, nor will I ever take one. Am I smarter than the average person? undoubtedly. But if my IQ is high then I fear I'll avoid doing useful things because they're beneath me. And if my IQ is not high, I fear I'll avoid doing things because they're above me. There is nothing useful that can come out of knowing your IQ, and as others have said, there's other things that are so much more important. Smart people are unusually good at deluding themselves, for example. They get so used to being right they start to believe if they can rationalize it, it's obviously true and anyone who disagrees isn't as smart. Having said that, I once made a girl angry in college because I asked the professor if grad school will actually be challenging because undergrad math for me was ridiculously easy. I think the best way to describe IQ is that it defines your potential (w/i a margin of error), but it's everything else that determines how effectively you use that IQ. A child with down syndrome will never win a nobel no matter how much they want to or how much effort they put into it. But having an IQ of 160+ will never guarantee you don't end up homeless in a ditch.
csa 4y ago
> IQ is effectively just a measure of test-taking skills. Omg, no. Good test taking skills will help someone score close to their theoretical maximum, but you can’t just train someone in test-taking skills and expect them to score at the top of the scale. I’ve trained people on test-taking skills to great effect, but there is a limit (often significant) on how much these skills can improve a score.

tgv 4y ago

IQ is a fair predictor of achievement in life, though. But there are other predictors too, such as parents' socio-economic status, or geographical location. We wouldn't claim these things measure intelligence, though. IQ is a proxy for intelligence, but one with great shortcomings. Unfortunately, there's no agreed upon definition of intelligence, let alone an operalization, and a way of measuring it. So IQ keeps being used.

mjburgess 4y ago
> Unfortunately, there's no agreed upon definition of intelligence, let alone an operalization, and a way of measuring it. So IQ keeps being used. Uhuh.... so this line it's crazy, right? It's really strange you concede that we have no theory and hence no way of verifying whether an IQ score is a valid and reliable measure of the thing we have no theory of... and yet, you say it is a "proxy for intelligence" The entire edifice of research which makes these claims is a product of a pseudoscientific use of statistics by amateur statisticians. The vast majority of the results you think exist in the literature are frauds, and often egregious ones at that. A human being is one of the most complex systems known to exist in nature, and the frauds in psychometrics are very far away from being up to the task of studying them. The greatest minds in applied sciences, matehmatics and statistics would largely throw up their hands at the job of measuring "intelligence" --- that some fools havent, is no good thing.
- tgv 4y ago
  We all "sort of" recognize intelligence, but it stops there. > The entire edifice of research which makes these claims is a product of a pseudoscientific use of statistics by amateur statisticians. That's bollocks. It's just a topic that's too complex. Unfortunately, everybody likes it and draws conclusions from it. Even the HR department. If you would say we should ignore IQ measurements in society, I would agree. But it's not pseudo-scientific use of statistics. Nor all done by amateurs. It might even be more scientific than string theory or dark matter.
  - mjburgess 4y ago
    Alas, it is. It's severely pseudoscientific. None have PhDs in statistics. They're all people who got into psychology and were given GUIs by the maths department. Their papers are a symptom of what happens when you press the "ANONVA" button and print the results. The entire field is an egregious testament to pseudo-intellectual pseudoscience. It's shocking. There's nothing there. It's unfathomably bad.
    - tgv 4y ago
      The statisticians I knew were only too happy to go along with some ANOVA. They're usually consulted at the start of the study, but rarely co-author. A lot of psychology is bad research. There's a reason there's a reproduction crisis. Statisticians are to blame for that, too. They've been teaching and using NHST, even though all arguments point away from it. But statisticians can't fix a bad design, because they rarely know the factors involved, and they can't fix errors in the methods, because they've got no idea about the subject matter. The statistician in your argument comes across as some kind of totem. > There's nothing there. There's something there, but since we don't know what, it's best to take it with a grain of salt. However, if even the bloody HR department looks at IQ scores, it gets a lot of attention. And one explanation is that it predicts better than nothing. If you have to choose between two candidates, and one has an IQ of 80 and the other of 120, you know who (on average) who fits best.
      - 4y ago

ilyt 4y ago

"test-taking" is too generic. Puzzle-solving would be more accurate. IQ doesn't do much for test that just require to remember a lot of stuff.

maximus-decimus 4y ago

But IQ wasn't invented to predict "life outcome", it was invented to find kid's academic potential. I'm still waiting for somebody to show me a prolific Math Phd with below average IQ. The only counter example I've heard people use it Feynman whose IQ was "only" 125 putting him smarter than "only" 95% of the population. It rubs me the wrong way when Hawkins says stuff like "IQ is for losers" but then boasts how his physics undergrad degree was so easy for him he was just bored out of his mind (it sure wasn't easy for me). It's the intelligence equivalent of a billionaire saying you just need to work hard to get there, as if that was ever gonna make a difference if you're an Amazon factory worker.

mjburgess 4y ago
It was invented as a clinical test of retardation, and the only correlation with IQ and objective independent measures of performance is in the region of very low scores. Note, that when carear-IQ correlations are brought up it is universally in those populations subject to an IQ test filter who are then shown to correlate with IQ scores (ie., those who take, eg., MCAT, LSAT, ...) --- which, if it isnt clear, is circular. Average and above IQ shows no correlation whatsoever with, eg., wealth.
P5fRxh5kUvp2th 4y ago
I've never taken an IQ test, nor will I ever take one. Am I smarter than the average person? undoubtedly. But if my IQ is high then I fear I'll avoid doing useful things because they're beneath me. And if my IQ is not high, I fear I'll avoid doing things because they're above me. There is nothing useful that can come out of knowing your IQ, and as others have said, there's other things that are so much more important. Smart people are unusually good at deluding themselves, for example. They get so used to being right they start to believe if they can rationalize it, it's obviously true and anyone who disagrees isn't as smart. Having said that, I once made a girl angry in college because I asked the professor if grad school will actually be challenging because undergrad math for me was ridiculously easy. I think the best way to describe IQ is that it defines your potential (w/i a margin of error), but it's everything else that determines how effectively you use that IQ. A child with down syndrome will never win a nobel no matter how much they want to or how much effort they put into it. But having an IQ of 160+ will never guarantee you don't end up homeless in a ditch.

csa 4y ago

> IQ is effectively just a measure of test-taking skills. Omg, no. Good test taking skills will help someone score close to their theoretical maximum, but you can’t just train someone in test-taking skills and expect them to score at the top of the scale. I’ve trained people on test-taking skills to great effect, but there is a limit (often significant) on how much these skills can improve a score.