3 ms·

IQ is a model. Models never perfectly match the thing they are modelling. In the case of IQ people seem to have forgotten whatever it was that they were trying

by mantap 7y ago

IQ is a model. Models never perfectly match the thing they are modelling. In the case of IQ people seem to have forgotten whatever it was that they were trying to model in the first place and it has been promoted way beyond its station to the very definition of intelligence. The result of this is organisations such as Mensa. To me, intelligence means one's capability to contribute to human endeavour, whether it be in mathematics, music, art, writing, or anything else that you can apply your mind to. For example, most people agree that Mozart was a very intelligent individual by virtue of his contributions to music. Sure, IQ means something, it's just that the thing it means is not intelligence. It doesn't capture the amazing breadth of human capability, and intelligence is not an objective measurable property.

oska 7y ago
I agree with your contempt for IQ. On this matter: > In the case of IQ people seem to have forgotten whatever it was that they were trying to model in the first place IQ was originally designed to determine mental development issues, i.e. sub-intelligence, so that those people could be helped. This is flagged by scores below 80. Ranking intelligence by scores above 80 is bunk. Nassim Taleb has written well on this subject [1], if you are interested. [1] https://medium.com/incerto/iq-is-largely-a-pseudoscientific-swindle-f131c101ba39 https://medium.com/incerto/iq-is-largely-a-pseudoscientific-...
- LMYahooTFY 7y ago
  ...so he's claiming people conducting statistical analysis for IQ tests don't account for people dying or suffering traumatic brain injury?
  - lliamander 7y ago
    No. Taleb is saying that IQ is only useful for detecting mental deficiencies. He claims that the while having a low IQ is generally predictive of low life success (income, etc.) having a high IQ is not very predictive of having high life success. As others have pointed out, the same can be said of height and success in professional basketball. If you come across a short person, you can practically guarantee they cannot make it in the NBA (i.e. height is a strong predictor). For someone who is very tall, You can't really know from that whether they are a successful pro basketball player (i.e. it is not a strong predictor). This issue here, in both height and IQ, is that people who have more of some capacity have more options. Someone who is smart enough to be a brain surgeon still has the option to be a dish washer. Someone who is only smart enough to be a dish washer does not have the option to be a brain surgeon.
    - oska 7y ago
      I understand the point you're making but you risk giving IQ credibility by comparing it to height. Height is a real, physical uni-dimensional measure that has a Gaussian distribution. IQ is a made-up measure that pretends to fit a multi-dimensional, non-physical 'thing' (intelligence) that we have a lot of trouble actually defining, onto a uni-dimensional scale with a modeled Gaussian distribution. It's a fraud. Or at least, in this grossly extended use it became a fraud. Testing people for falling below a certain threshold on a test was an ok diagnostic for mental deficiency. But even there it should have just been one tool among many, not a final arbiter of "general intelligence".
      - lliamander 7y ago
        > IQ is a made-up measure that pretends to fit a multi-dimensional, non-physical 'thing' (intelligence) that we have a lot of trouble actually defining, onto a uni-dimensional scale with a modeled Gaussian distribution. I think one of the points of intelligence research is that it seems to confirm intelligence (whatever it is) is largely a unified by a single dimension, even when that wasn't necessarily what was expected by the researchers. > It's a fraud. Or at least, in this grossly extended use it became a fraud. Hardly. IQ testing has been vigorously resisted for decades, and yet continues because of it's utility. But if you're going to take that stance, let me ask you: are you willing to throw out the rest of the social sciences? Because that's the only logical result. IQ research is the gold standard of the social sciences. IQ is fraudulent, then so is the rest. > Testing people for falling below a certain threshold on a test was an ok diagnostic for mental deficiency. It is also useful at the higher end. Not all high IQ people are successfully working in complex jobs, but I expect all people who are successfully working in complex jobs are high IQ.
        oska 7y ago
        > But if you're going to take that stance, let me ask you: are you willing to throw out the rest of the social sciences? Because that's the only logical result. IQ research is the gold standard of the social sciences. IQ is fraudulent, then so is the rest. No, the 'logical result' is throwing out psychology, which I'm more than happy to do. Would also throw out 90% of economics. Happy to keep descriptive sociology, descriptive anthropology, descriptive human geography, descriptive linguistics, etc.
- lliamander 7y ago
  Taleb's critique is pretty deep. This will take me a while to parse through. One thing he tackles that I think is worth digging into is his approach to the claim IQ is psychology's "best measure". First, I think he does the only logical thing you can do after you dismiss IQ, he dismissed the rest of psychology. However bad (or good) you think IQ, the rest of psychology is worse. Second, he makes the point that "no measure is worse than a bad measure". I'm not entirely convinced. I mean yes, if you have a measure that only predicts (for example) 30% of the variance in some phenomenon and you treat it as if it explains 100%, then you're going to have a bad time[1]. However, if you recognize the limitations and you combine that with other measures, then you may have a chance at making reasonable predictions. IQ may be the best psychology has to offer, but that doesn't mean it has to be used alone. And it's true that much of psychology is suffering from a reproducibility crisis, but that does not mean all parts of psychology are suffering equally. [1] I think there's an interesting connection between this idea and the idea of legibility as touched on in the book Seeing like a State.
lliamander 7y ago
> To me, intelligence means one's capability to contribute to human endeavour, whether it be in mathematics, music, art, writing, or anything else that you can apply your mind to. Do you not think that people who have such capabilities would also have a high IQ?
LegitShady 7y ago
>IQ is a model. Everything is a model, but IQ is literally the most statistically significant idea in psychology. Your strict resolution that IQ doesn't mean intelligence is just your own opinion not really backed by any fact, just an opinion that people are more complicated than numbers. It's true, people are more complicated than numbers, but nonetheless they can be accurately rank ordered and IQ, as you said, means something. It doesn't say anything about accomplishments, or quality of character. But it does say something about fluid intelligence. Nobody cares if it perfectly matches some arbitrary thing you decided it needs to match to count for you. The mozart comment doesn't even fit into the conversation. Nobody said anything about contributions to anything. Fluid intelligence is a measurable property, even if its not measuring it to your personal satisfaction. The question becomes why are you more of an authority than those mathematicians and psychologists who did the research? You're not. Your opinion doesn't invalidate theirs. It just says you didn't do the research.

by mantap 7y ago

oska 7y ago

I agree with your contempt for IQ. On this matter: > In the case of IQ people seem to have forgotten whatever it was that they were trying to model in the first place IQ was originally designed to determine mental development issues, i.e. sub-intelligence, so that those people could be helped. This is flagged by scores below 80. Ranking intelligence by scores above 80 is bunk. Nassim Taleb has written well on this subject [1], if you are interested. [1] https://medium.com/incerto/iq-is-largely-a-pseudoscientific-swindle-f131c101ba39 https://medium.com/incerto/iq-is-largely-a-pseudoscientific-...

LMYahooTFY 7y ago
...so he's claiming people conducting statistical analysis for IQ tests don't account for people dying or suffering traumatic brain injury?
- lliamander 7y ago
  No. Taleb is saying that IQ is only useful for detecting mental deficiencies. He claims that the while having a low IQ is generally predictive of low life success (income, etc.) having a high IQ is not very predictive of having high life success. As others have pointed out, the same can be said of height and success in professional basketball. If you come across a short person, you can practically guarantee they cannot make it in the NBA (i.e. height is a strong predictor). For someone who is very tall, You can't really know from that whether they are a successful pro basketball player (i.e. it is not a strong predictor). This issue here, in both height and IQ, is that people who have more of some capacity have more options. Someone who is smart enough to be a brain surgeon still has the option to be a dish washer. Someone who is only smart enough to be a dish washer does not have the option to be a brain surgeon.
  - oska 7y ago
    I understand the point you're making but you risk giving IQ credibility by comparing it to height. Height is a real, physical uni-dimensional measure that has a Gaussian distribution. IQ is a made-up measure that pretends to fit a multi-dimensional, non-physical 'thing' (intelligence) that we have a lot of trouble actually defining, onto a uni-dimensional scale with a modeled Gaussian distribution. It's a fraud. Or at least, in this grossly extended use it became a fraud. Testing people for falling below a certain threshold on a test was an ok diagnostic for mental deficiency. But even there it should have just been one tool among many, not a final arbiter of "general intelligence".
    - lliamander 7y ago
      > IQ is a made-up measure that pretends to fit a multi-dimensional, non-physical 'thing' (intelligence) that we have a lot of trouble actually defining, onto a uni-dimensional scale with a modeled Gaussian distribution. I think one of the points of intelligence research is that it seems to confirm intelligence (whatever it is) is largely a unified by a single dimension, even when that wasn't necessarily what was expected by the researchers. > It's a fraud. Or at least, in this grossly extended use it became a fraud. Hardly. IQ testing has been vigorously resisted for decades, and yet continues because of it's utility. But if you're going to take that stance, let me ask you: are you willing to throw out the rest of the social sciences? Because that's the only logical result. IQ research is the gold standard of the social sciences. IQ is fraudulent, then so is the rest. > Testing people for falling below a certain threshold on a test was an ok diagnostic for mental deficiency. It is also useful at the higher end. Not all high IQ people are successfully working in complex jobs, but I expect all people who are successfully working in complex jobs are high IQ.
      - oska 7y ago
        > But if you're going to take that stance, let me ask you: are you willing to throw out the rest of the social sciences? Because that's the only logical result. IQ research is the gold standard of the social sciences. IQ is fraudulent, then so is the rest. No, the 'logical result' is throwing out psychology, which I'm more than happy to do. Would also throw out 90% of economics. Happy to keep descriptive sociology, descriptive anthropology, descriptive human geography, descriptive linguistics, etc.
lliamander 7y ago
Taleb's critique is pretty deep. This will take me a while to parse through. One thing he tackles that I think is worth digging into is his approach to the claim IQ is psychology's "best measure". First, I think he does the only logical thing you can do after you dismiss IQ, he dismissed the rest of psychology. However bad (or good) you think IQ, the rest of psychology is worse. Second, he makes the point that "no measure is worse than a bad measure". I'm not entirely convinced. I mean yes, if you have a measure that only predicts (for example) 30% of the variance in some phenomenon and you treat it as if it explains 100%, then you're going to have a bad time[1]. However, if you recognize the limitations and you combine that with other measures, then you may have a chance at making reasonable predictions. IQ may be the best psychology has to offer, but that doesn't mean it has to be used alone. And it's true that much of psychology is suffering from a reproducibility crisis, but that does not mean all parts of psychology are suffering equally. [1] I think there's an interesting connection between this idea and the idea of legibility as touched on in the book Seeing like a State.

lliamander 7y ago

> To me, intelligence means one's capability to contribute to human endeavour, whether it be in mathematics, music, art, writing, or anything else that you can apply your mind to. Do you not think that people who have such capabilities would also have a high IQ?

LegitShady 7y ago

>IQ is a model. Everything is a model, but IQ is literally the most statistically significant idea in psychology. Your strict resolution that IQ doesn't mean intelligence is just your own opinion not really backed by any fact, just an opinion that people are more complicated than numbers. It's true, people are more complicated than numbers, but nonetheless they can be accurately rank ordered and IQ, as you said, means something. It doesn't say anything about accomplishments, or quality of character. But it does say something about fluid intelligence. Nobody cares if it perfectly matches some arbitrary thing you decided it needs to match to count for you. The mozart comment doesn't even fit into the conversation. Nobody said anything about contributions to anything. Fluid intelligence is a measurable property, even if its not measuring it to your personal satisfaction. The question becomes why are you more of an authority than those mathematicians and psychologists who did the research? You're not. Your opinion doesn't invalidate theirs. It just says you didn't do the research.