Note: An earlier version of this article contained speculative data about how scaled scores related to IQ on the original WAIS. I have since been provided with the correct numbers and so this article was substantially revised on March 16, 2016.
As I’ve discussed before, a commenter named Andrew informed of me this study, where six samples of seniors from the extremely prestigious Dartmouth (the 12th most selective university in America) averaged 1357 on the pre-1974 SAT. I previously estimated that 1357 before 1974 equaled IQ 133 (U.S. norms); 132 (U.S. white norms). However because many Ivy League students presumably took the SAT multiple times and only their best scores are counted, it’s reasonable to deduct the equivalent of 2 IQ points from these SAT derived IQ equivalents, reducing them to 131 (U.S. norms), 130 (U.S. white norms).
Assuming these students are typical of high SAT Americans, it is interesting to ask how much they regress to the mean on various subtests of the WAIS.
Averaging all six samples together, and then adjusting for the yearly Flynn effect from the 1950s through the 1970s (see page 240 of Are We Getting Smarter?) since the WAIS was normed circa 1953.5 but the students were tested circa 1971.5, then converting subtest scaled scores to IQ equivalents, in both U.S. norms and U.S. white norms (the 1953.5 norming of the WAIS included only whites), we get the following:
iq equivalent (u.s. norms) | iq equivalent (u.s. white norms) | estimated correlation with sat in the general u.s. population inferred from regression to the mean from SAT IQ 31 points above U.S. mean. | |
sat score | 131 | 130 | 31/31 = 1.0 |
wais information | 128.29 | 127.2 | 28.29/31 = 0.91 |
wais comprehension | 122.22 | 120.9 | 22.22/31 = 0.72 |
wais arithmetic | 120.37 | 119 | 20.37/31 = 0.66 |
wais similarities | 119.16 | 117.75 | 19.16/31 = 0.62 |
wais digit span | 117.37 | 115.9 | 17.37/31 = 0.56 |
wais vocabulary | 125.93 | 124.75 | 25.93/31 = 0.84 |
wais picture completion | 105.87 | 104 | 5.87/31 = 0.19 |
wais block design | 121.82 | 120.5 | 21.82/31 = 0.7 |
wais picture arrangement | 108.33 | 106.55 | 8.33/31 = 0.27 |
wais object assembly | 113.65 | 112.05 | 13.65/31 = 0.44 |
wais verbal scale | 126 | 125 | 26/31 = 0.84 |
wais performance scale | 116 | 114 | 16/31 = 0.52 |
wais full-scale | 123 | 122 | 23/31 = 0.74 |
Information and Vocabulary have the strongest correlation with SAT
Of all the individual subtests, Information correlates most with SAT scores, followed by Vocabulary. This makes sense because Information (a test of general knowledge), like the SAT. measures verbal and numerical acquired knowledge, and Vocabulary, like the verbal SAT, measures verbal acquired knowledge. Also, Information and Vocabulary are highly g loaded, and should correlate well with all tests.
It is interesting that two commenters on this blog with extremely high SAT scores have reported very high scores on these subtests. Black national merit finalist ruhkukah obtained his two best WAIS-IV scores on Information and Vocabulary (both at the 99.9 percentile, and even this might be an underestimate because this is the ceiling of these subtests).
Meanwhile commenter chartreuse, who scored 1560 on a version of the SAT, and perfect on the GRE (which is very similar to the SAT) notes that his highest Wechsler subtest score from childhood was on Vocabulary. He did not state his Wechsler Information score, but did report performing twice his chronological age on another general knowledge test from childhood.
wais picture arrangement
How is it not possible to score very well on that? Vocab is much harder . I remember many years ago doing a picture completion test and it seemed inconceivable how a reasonably competent couldn’t get a perfect score
these words also seem too easy
[image redacted by pumpkin person, feb 4, 2016]
The analogies section of the old SAT is much harder
I had to redact the vocab list you post because some of those words could still be used to test people. Many of those words seem easy but one of them is constantly misused so obtaining bonus points for precise understsnding is rare.When i see someone who uses that word correctly, it indicates extremely high verbal IQ
Can you please un redact this? I genuinely want to test myself/ see my understanding of these words.
Susan Carry. 34:30
She is wrong about integers. 37:30
He was not making an evolutionary argument.
You cannot have a fraction of a quantum.
Unless space is a continuum not discrete.
Cantor did not believe the infinitesimal existed.
Cantor believed in discrete space.
What do people with high visual and fluid intelligence but low verbal end up doing with their lives, if they don’t end up at places like dartmouth?
Good question
I assume they become artists, architects, pilots, engineers, chefs, actors etc. Martha Stewart is a good example of someone who got super rich off of non-verbal IQ (crafts, decorations, cooking etc). Rachael Ray is following in Martha’s footsteps, and claims to have a huge head, and is said to have an IQ in the 150s.
Is the modest correlation between math and PCI here due to the low standard of SAT math, or Dartmouth selection bias for verbal abilities, or is mathematical ability not well measured by the PCI in general?
*PRI not PCI
It’s hard to say because Ivy League students are selected based on composite SAT scores. If we had a sample that were selected just based on math SAT, they might have much higher performance IQ
Yes, PP, the Digit Symbol was used to calculate PIQ and FSIQ in the “old ‘ WAIS. with a range of scaled score 0-19.The sum of scaled scores for the test is 202 (FSIQ) =112 (VIQ) +90 (PIQ). Maximumpossible scaled score is 19 for all the subtests but Ar (17) PC (18) BD (17) PA and OA(both 18)
Thanks for the info. Can you email me the conversion table of sum of scaled scores to verbal IQ, performance IQ, and full scale IQ for the WAIS?
easiestquestion@hotmail,ca
[pumpkin person feb 7, 2016: the above email contains a typo. The correct email is easiestquestion@hotmail.ca ]
I want to see where the authors of this study went wrong, if they did:
Click to access 2532.pdf
Sure. Give me a little time
Thanks
Pumpkin – Jewish Manhattan and its High Verbal IQ connection, according to Lion:
https://lionoftheblogosphere.wordpress.com/2016/02/05/surprised-to-see-this-in-a-ny-times-article/comment-page-1/#comment-95994
Perhaps he is right. There are not many East Asian professionals working and living in Manhattan, when compared to other East Coast American cities.
I don’t understand. East asians tend to score normal in verbal iq, they tend to score very good in the verbal sector on Sat and others.
By psychometric parameters, east asians no have detectable lower verbal intelligence than other groups IF general intelligence and verbal intelligence are very mutually correlated among goym or non-jooische peuple.
I though the differential that made east asians weak in social network or communication professions is its lower % of narcisistics among them and not exactly by ” pure cognitive” reasons.
East Asians’ lack of narcissism – which is why Jorge Videla was ranting about Jews owning America. Our country rewards eloquent narcissists and intelligent sociopaths, not quiet, nice and polite people.
Verbal IQ would be useful in Manhattan because there are a lot of incredibly powerful jobs in finance and media where you manipulate symbols (words and numbers), as opposed to performance IQ where you actually have to solve practical problems in concrete physical reality.
check your email
I haven’t recieved it but its not your fault. There was a typo in the email i provided above. The correct email is:
easiestquestion@hotmail.ca
Sorry about that
Another reason why Ashkenazi Jews in America are successful. They are influential/entrenched in the most important cities. New York/Manhattan is a perfect example, and many of them control its real estate. But I’m not sure how this is connected with high verbal IQ.
I saw the typo and did it right first time. Are you sure that you don’t have any filter for dot ru domain? Ive just sent it again
Im gonna do it again from gmail in a minute
I see it in my junk mail folder. Thank you so much!
I expect some problems.
Firstly, you probably wanted the really-really real the WAIS-1955 manual in English. No, I don’t have it. But this is widely known here that the normative tables for The “Russian” WAIS were taken from the original WAIS without any change . So, my advice: don’t worry about this.
Secondly, all the links worked normaly, but, after placing them into a body of the letter, I see that 2of 3 don’t lead me to needed pages. If it happens to you too, Ive got some tricks to fix it
Is everething really ok?
Yes I was able to find all the tables I needed though it took some looking. I will have to update this post soon, since seeing the actual WAIS stats changes a few things.
Fascinating that they translated the American WAIS into Russian instead of just making a Russian WAIS. So many of the verbal items on the WAIS are rooted in English and American culture that it seems unfair to judge Russians by the same norms, even if questions are translated. But perhaps if they exclude Information and Vocab, and prorate for the missing subtests, it might work somewhat.
Ok, PP, great to know that it can help
So I see you updated this most. It seems it moved the full scale score substantially lower.
Yes. I had originally assumed that the sum of scaled scores had the same relationship to composite IQs on the WAIS as it did on the WAIS-R. That assumption was false.
Information 140
(40 * 0.91) = 136
Vocabulary 120
(20 * 0.84) = 116
Can you list the original non-corrected scores?
If i can find the study. The link to it is broken & i don’t recall the name of the paper
I’m an undergraduate at Harvard. We went over the WAIS-IV, live, in class (psychometrics) last week. The testing scenario was pretty informal, as we all were just yelling out answers to questions when they showed up, but I feel fairly confident in saying that at least one person in that class could achieve a 160+ on the WAIS without too much trouble (I have a specific person in mind). Although, we never went over the coding section. If I had to guess, I’d say that if we all took the WAIS-IV, the average IQ would indeed come out to about 130. Many of us were also given a version of the Wonderlic. We’ll go over our scores on that today I believe.
I’d be very interested in knowing how Harvard students perform on the wonderlic
How did you norm the intelligence tests on your website?
The norms on the matrix test seem about right but I think your verbal IQ test inflates scores by about 10 points (at least at the high end).
But you said your test’s FSIQ correlates at 0.95 with the WAIS-IV, so who knows.
Oh I recognize you. I took one of your IQ tests on your website a while ago. I didn’t know you commented here.
I scored a 27/30 verbal, equal to a 139 verbal score. Do you know if you penalize for guessing incorrectly?
If not, you should. The validity of the test would skyrocket if not already employed into the test’s construction.
pumpkin, iqtest.dk is famous for underestimating IQ, do you think it does too?