So you can look up how well it correlates with what things, and what it doesn't, and get an idea of what it is predictive of.
It isn't anywhere near perfect of course, you can't capture the whole of a person's cognitive ability in one number, but then it is only a "proxy" as you said.
Mensa is an association of moderately high-IQ people, and IQ correlates positively [0] with education and income, among other things.
[0] https://en.m.wikipedia.org/wiki/Intelligence_quotient#Social...
Specifically, when looking at people with IQ above 110, it seems like the correlations are still there at least for PhD level science. Most fields have average IQs above 130: http://www.religjournal.com/pdf/ijrr10001.pdf
One thing I didn't know initially is that IQ tests are normalized so the mean is 100 and the standard deviation is 10. If we assume IQ is normally distributed (not exactly true, but roughly accurate), then IQ of 110 puts you in the top 33% of the population. IQ of 130 puts you in the top 0.3%.
No, the standard deviation is 15 IQ points, not 10. 130 puts you in the top 2%. I know this because my kid scored in the < 0.02%.
On the other hand, IQ tests are highly correlated to how much you prepare for them, which is why rich families score higher. There was a tutor on Reddit for private schools admissions who said she had never had a student who didn’t surpass the entrance qualifications for a standardized test. It’s very easily gameable.
Guess the rest of us are potted plants.
Somebody better tell the "IQ measures innate intelligence" people.