alt Hacker NewsI would never ask for them since its so cringe. But SAT scores correlate to IQ at .81, and IQ is one of the few things that strongly correlates to knowledge work performance positively. There is probably a lot of alpha from knowing candidates SAT scores. Its more useful than knowing the college they went to.
Replies
tg180 • yesterday at 10:39 PM
[flagged]
doctorpangloss • yesterday at 10:35 PM
> and IQ is one of the few things that strongly correlates to knowledge work performance positively
i think you mean that it correlates to pay. nobody knows what you mean by "knowledge work performance." reviews of your peers also correlated with pay. often it is not the smartest person who is the most popular. so... do you see how you said something kind of meaningless?
➕ show 2 replies
That's r = 0.81 right? so R2 would then be ~.66. Meaning that 66% of the variation in IQ scores can be statistically explained by variation in SAT scores.
It's really high for psych stuff. If you even get r=0.5, you've got a great result there.
But it is important to note, I feel, that SAT maps to only about 2/3 parts of the IQ score, and IQ score is also a quite fuzzy measure here for things like knowledge work job performance.
I do agree though, you get quite a bang for your buck just reporting these numbers.
But, if you explicitly tie money and compensation to the SAT score, man, that is setting up some very perverse incentives around it. If it adopts widely to do so, then you're gonna get some really strange interaction effects there.