Tags
Commenter blubbalubba writes:
When will you do another study on the post 2016 SAT? There was a 60 points inflation. It would be great if you looked into this recent change.
For years I’ve complained that the SAT is normed on college bound students, but in order to convert it to IQ equivalents, we need to know how the general U.S. population would score. Well perhaps someone from the college board was reading because their website has now started publishing two types of norms:
Nationally Representative Sample Percentiles are derived
from a research study of U.S. students in grades 11 and
12 and are weighted to represent all U.S. students in those
grades, regardless of whether they typically take the SAT.
SAT User Percentiles are based on the actual scores of
students in the past three graduating classes who took
the current SAT during high school. These user percentiles
are reported on tests completed in August 2019 through
June 2020.
So for example, a combined score (verbal + math) of 1055, which makes you dead average (50th percentile) among the college bound elite who takes the SAT, would make you slightly above average (59th percentile) among the nationally representative sample.
One problem is even the nationally representative sample is a bit elite because many of the dullest Americans don’t make it to grade 11, let alone grade 12. so how do we adjust for this?
It turns out we don’t have to because only 6% of U.S. high school students now dropout, and even assuming none of this 6% would score high, the percentiles would only change by a fraction of a percent. For example, if you score in the top 1 out of 100 high school seniors, including the high school dropouts would only move you to 1 out of 106. A trivial difference (at least at the high end)
Verbal SAT (Evidence-based reading & writing)
On this section the median high school senior scores 510 so let this = verbal IQ 100. The 98th percentile is 720 so let this = verbal IQ 130. Thus:
Verbal IQ = 27 + 0.14(verbal SAT)
Math SAT
On this section the median high school senior scores 505 and the top 2% score 740, so let these scores equate to math IQs of 100 and 130 respectively. Thus:
Math IQ = 36 + 0.13(math SAT)
Combined SAT (verbal + math)
Overall the median high school senior scores 1055 and the top 2% score 1480, so let these scores equate to composite IQs of 100 and 130 respectively. Thus:
IQ = 26 + 0.07(Combined SAT)
Is the SAT an IQ test?
I have never found a study correlating the SAT with Wechsler IQ in the general U.S. population but I did find such a study correlating the WIAT (which also measures reading and math knowledge) with IQ in a very representative sample of U.S. 16-19 year-olds (n = 142). Reading comprehension and math reasoning correlated 0.71 and 0.82 with Wechsler full-scale IQ respectively (WAIS-III WMS-III Technical Manual page 85-86).
However there are several reasons for thinking the SAT may correlate somewhat less: 1) the enormous importance of the test for some creates huge differences in motivation, test-prep and cheating, and 2) people diagnosed with learning disabilities can take the test without time limits, 3) my research on the Wechsler IQs of Ivy Leagues students suggest they regress at least a third of the way to the mean from their SAT scores, suggesting a correlation no higher than 0.66.
Pumpkin Person 
all of these changes in entrance exams are motivated by the race gap in scores and thus the earlier the version the most g-loaded.
iirc asians actually scored lower than whites on the GRE verbal AND analytic sections. that surpsied me as the analytic section was (or seemed to be) a lot more like the math section than the verbal section, BUT as flynn showed, asian americans outperform academically (in terms of grades and highest level of credential) vs their IQ. so it could be the whites who took the GRE were more select.
another way to express idealism is…
humans are not exactly a stranger in a strange land.
that is, THINGS CAN’T KILL YOU!
and life is ultimately just a very unpleasant dream.
because God is AWE-FULL. what else could he/she/it be?
BMAC culture didn’t influence vedic culture as much as believed. In the rig veda saraswati river is mentioned and worshipped. Saraswati river is in india.
the original ode to mengele…One track on the album, a cover of the Soft Boys’ “I Wanna Destroy You”…by The Soft Boys.
Interesting. High IQ societies stopped accepting the post 1995 SAT for admittance. Why was that? Did they feel that the SAT ceased to be properly psychometric for some reason? My pre 1995 SAT was 1 in 5000. What floor do you think that would put for my IQ? Do you think I could still be cognitively subnormal?
Also, would pumpkin please detail or link to his research on IQ and Ivy League standardized test scores?
i’ll respond to this tomorrow, either by leaving you another comment or writing a brief article.
I first detailed the evidence here. The average Harvard undergrad wechsler IQ might be somewhat higher than I claimed because of sampling bias & imperfect corrections for the flynn effect, but not much higher:
https://pumpkinperson.com/2015/04/13/do-harvard-students-have-an-average-iq-of-122/
i always assumed mensa did it out of respect/fear of the college board. IQ had become controversial so the college board no longer wanted the SAT to be considered an IQ test. everyone assumes there was some scientific reason for mensa not accepting post-1995 scores but i think it was political. You got the lowest score out of 5000 people? In the SAT population or the general population? what was your score and how do you know it was the lowest 1 in 5000? That sounds impossible.
They all did. Supposedly after the recentering they don’t have faith it’s psychometric. But some still accept lsat and ACt however. They also dumbed down the difficulty of the questions to make less g loaded on both sections, I believe, but I was hoping you’d have the dope. No 1/5k high.
in this short video a guy makes fun of melo and rr.
LOL, Mugabe doesn’t know what physicalism is.
This is what I mean when I call it an outright conspiracy. FT may be owned by a Japanese company, but like investment banks, its not the shareholders that get all the value from owning a newspaper. I bet Bezos and Slim have less power to get a headline into their own newspapers as Mossad does. I’m not joking. I genuinely believe a deep stater with Mossad has more editorial control than the trophy owner of these papers.
Bezos might be the richest man in the world but he is NOT the most powerful man by any stretch of the imagination. Last year the saudis hacked his phone and published all his love correspondence with his mistress.
If Bezos was powerful it would never happen. Just like it never happens to Putin or Xi Jinping who are actually powerful people.
The west is a strange situation where the figureheads and the Route 1 conspiracy theorists ideas i.e. rich people are both less powerful than what I believe to be the true rulers of the West.
And yes, there are rulers of the West. The US millitary doesn’t invade random countries just to make money for boeing and lockheed shareholders.
You don’t think Putin could be hacked?
There is a hand at the Tiller. Someone controls the CIA and M16 like puppets. They didn’t invent Russia Russia by themselves.
Anyways these are my thoughts on the news all over the british papers today that russia hacked the Brexit referendum.
just ignore the news, like i imagine the more intelligent soviets and north koreans ignore their news.
what happens when stupid and evil people occupy the commanding heights?
then being poor, re-wilding, is a mark of superiority.
Anyone here in Mensa?
all high IQ societies are for autistic people with low IQs.
i myself belong to the David Ahenakew Society.
but i’m the only member.
Well we all know you aren’t.
mugabe is such an animal rights extremist he sees anti-animalism everywhere?
In the past few years, Hindus of the higher sort of castes have become even more obsessive than normal about protecting bovine rights, with vegetarian vigilante mobs lashing out violently at Untouchables, Muslims, and Christians for disrespecting cows. Cow Lives Matter!
well YEAH. THEY DO! THEY REALLY DO!
https://www.takimag.com/article/our-new-religion-of-race/
the supply curve is often or usually downward sloping because “fixed costs” become an ever greater fraction of costs as automation continues from machine age to information age.
and automation replacing people is the biggest argument for UBI.
these two facts are CONSISTENT.
at the extreme one could imagine that robots did everything that anyone might want done…but if all the robots were owned by one human being…how would the other humans pay for their labor?
higher and higher fixed costs = more free money actually increases production = higher standard of living for non-robots
Your ceiling for the SAT (M+V) is just short of 140.
Given that you have taken 2sd as reference point, your formula can’t be right because SAT goes up to 3.3 sd among test takers (wich is higher for general population as you said).
You should get a ceiling around 152.
where are you getting the 3.3 SD figure?
I agree with Bruno. If you look at the nationally representative percentiles, the 99th percentile starts at 1450, not 1560.
https://s29981.pcdn.co/wp-content/uploads/2018/11/pdf_understanding-sat-scores.pdf
I remember there was 650 top scorers out of 1 600 000 test takers that’s 1 in 2400. In IQ it would be around 150 or 3.3 sd.
https://blog.prepscholar.com/sat-percentiles-high-precision-2016
Then if you add the 3 IQ points for an average at 59th percentile, you get a top score at 153, or 3.53 sd.
Interesting. I wonder why simple linear extrapolation didn’t work since they presumably normalized the distribution on a fairly recent sample.
Although your source did include this caveat:
The table above is from the old SAT, with imputed new SAT scores. This means that current SAT scores are not directly from the current SAT; they are calculated from how old SAT scores convert to new SAT scores. This is good enough if you want a rough mapping between the two. Unfortunately, the College Board no longer releases exact numbers of test takers for each score, so we can’t update this table for 2020 (though the percentiles shouldn’t have changed much anyway!).
And the 2400 had probably a 156 ceiling wich is quite good. Must have been one of the highest ceiling IQ test. Almost like Wechsler. I didn’t realize that b4. That goes with Mugabe and LoB theory about SAT being the best IQ test.
And for ultra-high Math IQ, my theory is it’s the international math Olympiad, if they can reach 10% of the math talent, with a ceiling at 172 and 178 if they reach 100% of the existing math Talent. It’s probably around 175 (one third of the existing talent).
A perfect score at IMO (42) should qualify for Mega Society. Someone should tell them they have an available secure test at 1 in a million (even a bit higher).
Hey pp what do you think about Terence Tao’s head form? He struck me as a rather long headed person in this comparison:
yes he does have a long head.
Would this work if I studied for the SAT (like, reviewing every single answer I got wrong), if I average all of my scores?
Your formula here roughly correlates to what Emil Kirkegaard has posted on the SAT to IQ conversion. I took the SAT in high school in the early 2000s and got a slightly lower verbal score than the practice test I just took on the khan academy site being curious if I’ve had any mental decline over the past 20 years, which makes me think it is currently slightly easier (I’m pretty sure my brain worked a bit quicker back in the day, if slightly less thorough…).
Anyway, good seeing your thoughts about it.
Did you account for the average IQ of the test trackers being above average. The average IQ in the United States is 98. Around 15% of high school students drop out and not all students take the SAT or ACT unless they are in a state that either mandates it for graduation or they plan on attending university. This means your equation is showing people who have IQ’s in the mid 120s as being much less intelligent at around 110.
If you took this argument to its extreme it would be impossible to get a gage of the LSAT to IQ if you did not account for this difference. The average LSAT taker probably has around a 120 IQ being a subset of university graduates planning on going to law school. The average university graduate is around 115. So if you did not account for the test pool being from a smaller sub set of the population you would show test takers with a 120 IQ (top 10%) as being average at 100.
This is all explained
old Drake is the best Drake
Interesting. Was there really that much of a drop in IQ correlations compared to the pre-2016 SAT? If so, why do you think that is the case?
Overall no, but probably at the high end because they reduced the test’s ceiling so much.
Well, a 1400 on the old pre-2016 SAT corresponds to an IQ of 23.835 + 0.081(1400) or a 137 whereas a 1400 on the new SAT corresponds to a IQ of 26 + 0.07(1400) or a 124. That’s a huge drop, although 1400 might not be on the high end in your book.
Furthermore, is it possible for you to do a post regarding the conversion of post-2016 PSAT scores to IQ? It seems to have a much higher ceiling and it is less common for students to prepare for it than the SAT. I have a PDF of the percentiles here and :
Click to access psat-nmsqt-understanding-scores.pdf
why are you obsessed?
Seems like it would be a bit redundant. The nationally representative percentiles can be readily converted to IQ.
You mean the nationally representative percentiles for the PSAT?
That’s right
Why would the PSAT have a higher ceiling than the SAT, which seems to be the case here?
Because the SAT’s ceiling was probably deliberately lowered to make college affirmative action less obvious. The PSAT by contrast is not really used for college admission.
i got a PSAT score in the high 180s on the my first attempt and a low 190s on my second attempt.
i recall those being very high scores.
Hi Pumpkin, if we are interested in a basic conversion of the post-1995 pre-2016 scores (mine was from 2003), should we use your old post here https://brainsize.wordpress.com/2014/06/12/converting-post-1995-sat-scores-into-iq/?
I scored a 1550 in 2003 (780 M, 770 V) so I’m curious about if there was a different conversion then, given the post-2016 inflation you talk about.
You took what I like to call the Generation Y SAT (taken from April 1995 to March 2016).
Instead of the link you provided, I would use the formula at the bottom of this post:
It’s based on more data.
Great, got it, thanks pumpkin! That’s interesting that the ceiling was that high.