To mathematically estimate the IQs of these people, I used publicly available data such as scholastic achievement scores, ethnicity, GPA, wealth, self-reported IQ scores, and so on. Generally, I try to avoid using too many predictors as these variables will covary for other reasons besides intelligence, so continuous regression will result in mistakes.
This method of estimating IQ–using high school SAT scores–will *always* yield an underestimate for intellectual outliers–especially intellectual outliers with assymetric cognitive profiles. Once you get past IQ 145, you have to incorporate qualitative evidence into the analysis. This may not seem as “objective” as regressing on high school test scores, but it’s necessary nonetheless.
>1540 on his SATs (V800; M740) and a 30 on his MCAT.
I think this increases it by about 5 points or so. Not a big difference, but I appreciate the information.
>This method of estimating IQ–using high school SAT scores–will *always* yield an underestimate for intellectual outliers
Not really. I actually messed up the original formula - I believe outliers will regress to the mean by even more than what I originally envisioned, but I'm going to go over that in the second update with (hopefully) about 100 figures and a better methodology.
Yes, I understand that people like Aella and Scott do not come from a representative population, but the information they disclose doesn't either. Humans are much more likely to disclose favorable rather than unfavorable information, and it is difficult to factor this in when doing the analysis.
>Once you get past IQ 145, you have to incorporate qualitative evidence into the analysis
Not necessarily. I don't think qualitative evidence is going to be a very reliable way of estimating IQs, unless it is for identifying very obvious outliers (e.g. Von Neumman). Looking at the Emil survey, people's subjective estimations of the IQs of public figures are very discordant on an individual level, so I don't see why the same shouldn't apply to me.
Wordpress ate my comment when I tried to respond to your post estimating Scott Alexander's IQ, so I will just post my comment here:
This whole analysis is off, but the second paragraph is verifiably wrong. You take Scott’s self-deprecation too seriously.
He scored a 1540 on his SATs (V800; M740) and a 30 on his MCAT.
Source: https://www.reddit.com/r/slatestarcodex/comments/8e2838/ama_request_with_scott/
This method of estimating IQ–using high school SAT scores–will *always* yield an underestimate for intellectual outliers–especially intellectual outliers with assymetric cognitive profiles. Once you get past IQ 145, you have to incorporate qualitative evidence into the analysis. This may not seem as “objective” as regressing on high school test scores, but it’s necessary nonetheless.
>1540 on his SATs (V800; M740) and a 30 on his MCAT.
I think this increases it by about 5 points or so. Not a big difference, but I appreciate the information.
>This method of estimating IQ–using high school SAT scores–will *always* yield an underestimate for intellectual outliers
Not really. I actually messed up the original formula - I believe outliers will regress to the mean by even more than what I originally envisioned, but I'm going to go over that in the second update with (hopefully) about 100 figures and a better methodology.
Yes, I understand that people like Aella and Scott do not come from a representative population, but the information they disclose doesn't either. Humans are much more likely to disclose favorable rather than unfavorable information, and it is difficult to factor this in when doing the analysis.
>Once you get past IQ 145, you have to incorporate qualitative evidence into the analysis
Not necessarily. I don't think qualitative evidence is going to be a very reliable way of estimating IQs, unless it is for identifying very obvious outliers (e.g. Von Neumman). Looking at the Emil survey, people's subjective estimations of the IQs of public figures are very discordant on an individual level, so I don't see why the same shouldn't apply to me.
i'll take the under on Aella, Kanye, Magnus and Yarvin. I'll take the over on Yud, Light, and probably Scott.