On my other blog I wrote the following:
The SAT is said to correlate about as well with IQ as two different IQ tests correlate with one another. In his book Real Education (pg 69-70), Charles Murray claims that only 17 year olds capable of getting an 1180+ on the post 1995 SAT (Critical Reading + Math) are true college material. Although 35% of American 17 year olds take the SAT, he estimates that only 10% of 17 year would score 1180+ if all of them took it.. In other words, 1180+ is equivalent to an IQ of 120+.
In his book Coming Apart (pg 375) he estimates an SAT score of 1400 is equivalent to an IQ of 135. From these two data points, we can create the following formula for converting SAT into IQ equivalents:
IQ = 39.545 + 0.068(SAT score)
Note, this formula only applies to the post 1995 SAT. Before 1995 the SAT was much harder.
This formula works well for individuals, but should never be used to estimate the median IQ of a college. This is because elite colleges select for SATs, so the estimated median IQ will be artificially inflated by a selection bias.
Now when I plug in a perfect SAT score of 1600 into this formula, I get an IQ of 148 (U.S. norms). The problem is, I just read that 453 students scored 1600 on the new SAT in 1996-7 (out of approximately 3,500,000 17 year olds in the United States). That means 1600 equates to an IQ of 154 (U.S. norms).
So drawing a line of best fit through this data point, in addition to the two I got from Charles Murray, gives the following revised formula:
IQ equivalent = 23.835 + 0.081(new SAT score)
Pumpkin Person 
Seems to be quite similar to http://www.iqcomparisonsite.com/SATIQ.aspx
It’s always a good sign when different independent investigations arrive at the same conclusions
Indeed, PP
interesting, the site may be protected from linking, sorry
That’s a pretty high threshold according to Murray, but others have said college is ‘dumbed down’, so it can’t be both. I suppose the threshold would apply to STEM an not other majors?
Murray makes a distinction between dumbed down colleges & genuine college reading. In other words Murray feels the vast majority of college grads have not recieved a genuine college education in any field (STEM or arts)
Has there ever been a time when the majority of college students were IQ 120+ (pre-dumbing down)? That seems like a really high threshold. Nowadays only Ivy League-level universities have average IQs that high (basically, UC Berkeley and higher).
Given how rare it was to attend university in the early 20th century, and given how high the correlation was between IQ and years of education, I would expect the average IQ was indeed 120+, but I also suspect there used to be much less IQ difference between universities than there is today.
BTW, I got the UC Berkeley average IQ = 120 from the “g factor”, but of course according to the formula above it would be much higher. If one does a straight conversion from SAT to IQ with no regression, many schools nowadays would surpass the 120 IQ threshold.
Here’s an interesting side note note that corroborates your formula:
The average SAT for students coming out of the Davidson Academy, which only admits students who’ve demonstrated an IQ at or above 145, is 2230 on the three test scale, or approximately 1500 on the M+R scale (I use 1500 rather than 1490 because Math and Reading scores tend to be slightly higher than Writing ones). This, when put through your formula, equates to an IQ of 145.335. Almost perfect.
A similar degree of accuracy can be witnessed based off of Pine View School in Sarasota’s SAT – IQ correlation. The IQ required for admittance at the school is 130, and the average SAT (2008-09) was about 1320 on the R+M scale. This equals 130.755 on your formula. Uncanny.
I’ve looked at the Davidson Academy website before: http://www.davidsonacademy.unr.edu/Articles.aspx?ArticleID=144
I scored 600 verbal, 570 math in the 7th grade, so I was nowhere close to making it in 😦
I took it for the Duke TIP program.
Are you homeless now?
Your scores may not be Davidson level, but those are nonetheless very impressive scores for a seventh grader. Most of my classmates score well below that even in the 11th grade.
In retrospect, I wish I’d the SAT in 7th grade. My current score stands identical to the average at Davidson, making me wonder if I could’ve gotten in. My gut is that I’d have been able to easily make the verbal but not the math cutoff. 660 is a lot of math for a seventh grader. I’d be surprised if many of the entering seventh graders at Davidson can actually perform at that level.
the math on the SAT is elementary. I don’t think you need to know anything above algebra 1 to get a high score.
Not true, actually. Not even close. The math section is chock-full of geometry, which isn’t studied until after Algebra 1. The test also includes functions and graphing problems that are more Algebra 2 than Algebra 1 material.
But a lot of those students are selected based on SAT scores (taken at a younger age), so the high correlation may not be as uncanny as it seems
how many iterations has it been?
i took it at least 2 iterations before now.
the purpose of each iteration has been to make it LESS and LESS a measure of ability.
check out steve shoe’s latest posts.
PARADIGMATIC CASES OF WHAT A SHITTY COUNTRY AMERICA-STAN IS.
the guy’s parents aren’t even americans. he’s a chinese SUPREMACIST and a chink.
he exemplifies the following phenomenon:
the US DEPENDS on immigrants BECAUSE immigrants are GULLIBLE.
THEY’RE EASILY FOOLED.
THE SAME GOES FOR PROLE WHITES LIKE CHARLES MURRAY AND PEGGY NOONAN…
EXCEPT PROLE WHITES HAVE BEEN SCREWED BY THE LAST 35 YEARS IN…
AMERICA-STAN…
BUT ARE SO FUCKING RETARDED THEY STILL VOTE GOP.
JS WILL KNOW.
AMERICA’S RULING CLASS IS THE LOWER CLASS OF NON-SHITTY COUNTRIES.
But Hsu is at least pretty intelligent. When he talks technical you can tell he knows what he is talking about. Like, check out this talk
there are right wing whites who get it and arent proles / slaves. Like Jim Goad
shoe does have a facility with mathematics per se to some extent, though he’s made obvious mathematical errors on his blog.
so-called “higher mathematics” is too often over-awing for those without a background in mathematics.
believe me, however sophisticated the mathematical model it’s all shit if the model maker makes incorrect or simplistic assumptions…as does shoe.
he again exemplifies a general trait of chinamen.
they can be trained to do anything…except THINK.
shoe’s talk betrays egregious conceptual errors which i will elaborate anon.
Hsu makes no unwarranted assumptions. When he talks about additivity or anything, that’s not an assumption, that’s what the data shows. His mathematical models are robust.
About additivity:
http://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1000008
Both citizens and recent arrivals of the United Prole States are summed up in one sentence: Meriproles are striving boors.
Immigrants and prole whites share the same value of material worship and wasteful junk hoarding.
That puts me around 98.
I was 19 when I took it.
I hear that SAT is score relates to parental income. My mom had 3 kids and was divorced when I was 6 and was on welfare. My mom was 30 when she had me and my Grandmother was 41 when she had my mom. My grandmother saved money to give each of her 5 kids $100,000 when she died in 2005. I got $10,000 in 2014. I now have a house $74,000 mortgage. I do not know about my dad except he was in the army and has lost all his teeth. My brother spent $6,000 on root canals. I still have all my teeth. I went to a magnet middle school in the science one. I was in foster care between 14-16. I was held back a year from 3rd grade to 2nd grade because I was home schooled so when I went to school it was the first time I had to take a test. My mom taught us to read with Hooked on phonics. Before my grandmother got alzheimer’s she bought us lots of toys and books. Alzheimers was why my mom had to send the three of us to school. We lived on a square mile ranch property in the desert. My grandmother bought a $40,000 trailer for my mom to live in there. She had 5 properties this big. She sold goat milk from one and had a hotel. Both were in Albuquerque.
I am taking the compass test this friday. FASFA says I can get $5,775 in pell grants.
I need to take basic math and english before classes for a bachelor’s in CS.
I made this in Java:
http://pastie.org/10549279
Some online people helped me with the threading and synchronization but I did everything else.
Someone explained how Jungian functions actual work with 4 dichotomies that don’t match my dichotomies but I still think mine mean something?
Ni – Convergence
Ne – Association
Si – Immersion
Se – Sensuality
Fi – Will Power
Fe – Selflessness
Ti – Delineation
Te – Construction
It does seem that because of deep learning some organizations can now read minds / intentionality from mass data mining. Ripples of Novelty. Deviation from normal behavior. I would not be surprised if I am monitored. Windows ten made my java program run faster when cortana asked for feedback on my windows ten experience and I said java ran slow. One’s online presence can send echos. Like the echos of The Force. That was a good video game.
Good luck with college Animekitty!
By the way, I think you’re good example of why schizotypy should be added to the HEXACO/Big 5 personality model. It’ll better describe unique individuals such as yourself. 🙂
Is the fifth child usually at a disadvantage when it comes to intelligence. Having a child at age 41 and being the fifth one would have some effect wouldn’t it?
My uncle seems to be fine without schizotypy and he is 78 and a country singer.
According to Jayman, no effect of birth order on IQ:
I took the Compass test today
I don’t know all the statistics but I am 28 years old and I think this is the new test for 2015. I did this test on the computer in the state of New Mexico. I finished it in three hours 3:01:07
Math Placement
Pre Algebra Score 65 – 00:15:45
Algebra Score 31 – 00:24:12
Reading Placement
Score 92 – 00:45:10
Writing Skills Placement
Score 85 – 01:20:29
http://singularityhub.com/2015/11/13/technology-alone-is-no-miracle-drug/?utm_content=bufferd6ca3&utm_medium=social&utm_source=plus.google.com&utm_campaign=buffer
What impact would this have on the average /real/ IQ of students at say, Harvard?
Given that in the past you have predicted that students at harvard would have an average IQ of 135 based on a regression, and then later an IQ of 128 or below based on empirical data, would this increase your estimate?
I originally estimated the official IQ of Harvard students based on the correlation between IQ and academic success, on the very rough assumption that entering Harvard students were roughly the 1600 best U.S. students of their birth cohort.
I did not estimate the official IQ of Harvard students based on their SAT scores because I was not certain of the correlation between official IQ and SAT scores…instead I tried to estimate the correlation between official IQ scores and SAT scores from the degree of regression to the mean Harvard students show when they take the former.
All well and good but you’re not answering my question, what impact, in your view does this have on at least your estimate of the average IQ at Harvard?
Do you retain the point of view that their IQ’s are roughly between the mid 120’s to 130’s or based on the above data would you put them higher?
Would you say that your rough assumption holds or not?
Furthermore, I have another question and that is, given that students at Pine View have an average SAT score of roughly 1320, and an IQ of 130 is required to enter, would you not say that that’s fairly good evidence of a very strong correlation between IQ and SAT, moreso than previously thought?
All well and good but you’re not answering my question, what impact, in your view does this have on at least your estimate of the average IQ at Harvard?
I currently see no reason to revise my estimate of the official IQ of Harvard (as opposed to the SAT IQ equivalent). Instead I see reason to revise downward my estimate of the correlation between SATs and official IQ scores.
Do you retain the point of view that their IQ’s are roughly between the mid 120’s to 130’s or based on the above data would you put them higher?
I tentatively retain that view, yes.
Would you say that your rough assumption holds or not?
The rough assumption that Harvard students are roughly the 1600 best students of their birth cohort? No, it probably doesn’t hold.
Furthermore, I have another question and that is, given that students at Pine View have an average SAT score of roughly 1320, and an IQ of 130 is required to enter, would you not say that that’s fairly good evidence of a very strong correlation between IQ and SAT, moreso than previously thought?
No because a lot of them probably entered based on SAT scores (taken at younger ages), not official IQ scores. Further, whatever regression to the mean that might have occurred among those who did enter via official IQ scores might have been masked by the quality of Pine View education.
I’ve had a think pump, and given your stuff and interesting analysis, it seems to me that Mensa has let, based on their SAT Scores a bunch of mid IQ 120’s individuals into their club.
What are your thoughts on this, if I’m wrong why am I wrong?
You’re wrong because you’re confusing the much harder old SAT (used for Mensa admission) with the new post-1995 SAT discussed in my post
Irrelevant, Frey and Detterman used in their study the 1979 probability survey and came to the conclusion that when adjusting for non-linearity the correlation between old SAT and IQ is .857. This shows even the OLD SAT would show that even back in the day as it were, the SAT Dramatically over-estimates IQ apparently.
I also find it a little silly to suppose that, even with minority Quotas, the average IQ at Harvard, based on an average SAT of 1390 back then, was 144, and then dropped DRAMATICALLY to as low as 122!
As an anecdotal example also, Al Gore, having an IQ of 134 tested during his school years managed to achieve an 1355 on the pre 1974 scale which I am told is different even than the post 1974 Scale and more different still from the post 1995 Scale, graduated Cum Laude from Harvard.
If he represents the average Harvard student of his caliber which he may or may not do, but using him as a heuristic, am I to believe that the IQ at Harvard for whatever reason did a dramatic increase, then decrease?
I find this all remarkably hard to believe.
The Harvard IQ drops when they move from a test they were selected on (SAT) to a neutral test (WAIS). Regression to the mean
Of course Mensa has let many people with sub-130 IQs in. This is impossible to avoid because so many different tests are accepted, all of which correlate imperfectly with each other.
You could define IQ as being the highest possible score that someone could score on any of the accepted tests, then the rarity of a 130 IQ goes down.
I say all of this, because if you apply their formula, an SAT of 1250, taken to be equivalent to 130 IQ more or less, would be closer to 125.71 following their assessment.
Again, if I’ve misapprehended and indeed the average intelligence at Harvard has dramatically changed, feel free to point out why.
Well, Pump, hear me out:
Since the college board uses the SAT to select for aptitude for college generally, there IS no neutral SAT as there is a neutral WAIS.
Surely then, almost everybody that takes the SAT will regress on their scores when the WAIS is used to examine them. That is to say, because they are being selected on the basis of their scores.
Whether you got into Harvard or a state university, generally speaking you were being selected by the SAT, as a result we should expect to see the same regression everywhere.
Thankyou for suffering my questions so far.
Murray seems like a bit of a retard. I once read an academic paper of his that actually cited this page as a source:
http://www.statisticbrain.com/iq-estimates-by-intended-college-major/
Considering the WAIS-IV data that the average IQ of someone with at least a few years of college is only about 102, how can the average IQ of the SAT taker be 106 as your formula suggests?
Because not everyone who goes to college takes the SAT. Taking the SAT implies that not only do you plan to go to college, but you plan to go to a relatively competitive college that requires test scores.
I suppose that the “average college” may have admission standards so low that they bring the average college IQ down significantly
It appears that Pine View in Sarasota strictly uses official IQ tests for admissions.
“On the WISC-V the Full Scale Intelligence Quotient, General Ability Index, and Nonverbal Index are the only Index Scores that will be considered.”
http://www.sarasotacountyschools.net/schools/pineview/interior.aspx?id=66010
Under “Admissions”
http://teachersites.schoolworld.com/webpages/LLyons/
Surprised nobody checked this as soon as it was mentioned.
Do these idiot schools not realize how easy it is for parents to coach their kids on the WISC? The WISC is meant to be taken cold, and that doesn’t happen if it’s used as an admission test.
Do you have a formula for the ACT? I’ve read that the Math and English sections are the only ones that matter.
If not, what are your thoughts regarding the that test and IQ?
I don’t think Pumpkin has created a formula, but there has been a reported 0.77 correlation:
Click to access koening2008.pdf
There is a chart converting ACT scores to IQ scores:
http://www.assessmentpsychology.com/iq.htm
Pingback: IQ Roundup | greyenlightenment.com
Pumpkin, what about accounting for if you study for the SAT?