The below table is based on the first 29 people to take the Pieces test (excluding likely repeat attempts and people who spent less than 2 minutes). Since the totality of evidence now suggests that relative to their non-clinical general U.S. population, my sample has a Wechsler full-scale Z score of 1.93 (SD = 1.33) this distribution was then adjusted by the 0.6 correlation (no contamination) between WAIS-III full-scale IQ and Object Assembly (similar to Pieces) and then equated to the sample distribution to estimate the general population Z scores.
Out of all the subtests I have added to the PAIS, perhaps none were harder to construct than the multiple choice test Pieces. I wanted at least one subtest that measured pure spatial reasoning. I had already added cubes, however cubes is kind of a hybrid of spatial and working memory. Pieces was inspired by the Object Assembly subtest in the early versions of the Wechsler, which itself was inspired by the feature profile test used to screen out low IQ immigrants at Ellis island. However unlike these tests, you don’t actually put anything together in Pieces, you just answer 21 questions in 20 minutes about how you would do so which makes the test much more difficult and reliable though perhaps less fun and clinically informative.
Is the above picture not just the coolest thing you’ve ever seen in your life? I love old school psychometrics.
A much more detailed cardboard version of this profile puzzle would appear on the WAIS-III, WAIS-R, original WAIS and the granddaddy of all Wechsler scales: the ancient WBI, reminding us once again that Wechsler, who ironically was an immigrant himself, invented almost nothing. His genius was not creation, but curation. His tests became so popular because he did a brilliant job selecting his original 11 subtests and he did so based not exclusively on psychometrical criteria, but on his own intuition of what it meant to be an intelligent adult in 20th century America, which he defined as one who knows and understands the World around them, and has the resourcefulness to cope with its challenges. For Wechsler, clinical psychology was as much art as science and these old school hands on performance tests really measured if you could adapt to your environment in a creative way.
But their Achilles’ heel was low reliability which is probably caused by the luck factor of randomly fitting pieces together. It was also inconvenient to psychologists to lay out all the puzzle pieces in just the right position, so shortly after Wechsler died, newer WAIS revisions discontinued the subtest.
Using data from the first 55 unique individuals to take the sequence addition test, a cumulative frequency table was made:
This allowed me to convert raw scores on the test percentile equivalents and these percentiles were then assigned normalized Z scores (the Z scores they would equate to in a perfectly normal curve) relative to the test taking population, not the general population.
Raw
Percentile (%)
Z-Score (relative to the test-taking population NOT the general (U.S.) population
0
1.8
−2.08
1
4.5
−1.69
2
10.9
−1.23
3
20.9
−0.81
4
29.0
−0.55
5
36.0
−0.36
6
44.0
−0.15
7
59.0
+0.23
8
83.0
+0.95
9
96.0
+1.75
10
99.0
+2.33
The next question is how self-selected is the test taking population. On a scale where Americans average IQ 100 (SD = 15), those test takers who reported Wechsler IQs had a mean of 127 (SD = 24) suggesting they are much brighter and much more variable than the general U.S. population. There is no reason to think the test takers who reported Wechsler IQs are brighter than the test takers in general; in fact the latter scored higher on the sequence addition test (mean 5.65 SD = 2.51 vs mean 4.2 SD = 3.19) though given the small sample size of those with self-reported Wechsler scores (n=5) it’s not statistically significant.
Now when Ron Hoeflin normed the Mega Test, he used equipercentile equating, meaning that even though the two tests are imperfectly correlated, he assumed that the percentile distribution of the two tests would perfectly match. So if the highest SAT score among his Mega takers was one in a million, then the highest Mega score among the Mega takers was one in a million, even if they were not the same person.
Hoeflin’s approach makes sense IF one assumes self-selection for taking the Mega Test correlates about as well with Mega scores as it does with SAT scores, however I highly doubt my test correlates about as well with self-selection as the Wechsler does, and the reason is a comprehensive test like the Wechsler is going to be much more reliable and valid than my brief test is, and thus correlate better with most external criteria. So assuming my sample averages +1.8 SD on the Wechsler, I’d expect them to average (0.58)(1.8 SD) = 1.04 on sequence addition. Why 0.58? Because before contamination, that’s the correlation between WAIS-R full-scale IQ and Digit Span (the subtest most similar to sequence addition). Also, if my sample has a Wechsler SD that is 160% the U.S. SD, on this test they should be 160(0.58) = 93% of the U.S. SD.
Armed with these statistics, I transformed the normalized Z scores calculated with reference to the test taking population to Z scores with reference to the U.S. population.
U.S. population Z score = (test taker Z score)(0.93) + 1.04
Raw score on sequence addition
Estimated Z score relative to the U.S. population
0
−0.8944
1
−0.5317
2
−0.1039
3
+0.2867
4
+0.5285
5
+0.7052
6
+0.9005
7
+1.2539
8
+1.9235
9
+2.6675
10
+3.2069
I deliberately did not convert these scores to IQ equivalents because people shouldn’t get the impression that a simple test like this measures overall IQ but if I had, I would have done so by multiplying them by 15 and adding 100.
So as we march towards the end of the PAIS (Pumpkin’s Adult Intelligence Scale) we now have a subtest called sequence addition (aka Coordination). To register you need only provide a first and last name (real or fake).
I have a hobby of trying to guess IQs from a person’s most salient bio-demographic cognitive correlates (head size, money, religion etc) and commenter Teffec suggested I do this for the late great Charlie Kirk given his impressive height and influence. But influence is a bit amorphous so I decided to use his impressive occupation and height instead.
Although reports vary, according to Grok, Charlie Kirk is most commonly listed as 6’3.5″ which according to Grok, would make him 2.23 standard deviations above the average for American men of his generation (99th percentile). To put this in perspective, I had Grok list the 10 most successful talk show hosts of all time and provide their heights as Z scores relative to Americans of their sex and generation and the mean was 0.856 with an SD of 0.9. So relative to other elite talk show hosts, Kirk’s age and sex adjusted height is +1.7 SD (95th percentile).
Now in unrestricted samples the correlation between IQ and height is 0.25, but because talk show hosts show less cognitive variance than Americans as a whole, the expected correlation would be only 0.18.
127 (U.S. norms) 125 (U.S. white norms) (smarter than 96% of his generation and a bit smarter than the average super successful talk show host)
The standard error of the estimate is about 10.
Psychometric confirmation
I could find no publicly available data on how Kirk scored on the SAT/ACT but I may have found something about as good, which is hours of Kirk debating students at the most elite UK schools. After watching Kirk take questions from many young people at Cambridge and Oxford, I found five scenarios that were actual debates where there was a clear winner. Of these five, Kirk won three and lost two suggesting he’s smarter than 60% of Oxford/Cambridge students. Since we know these students have an average IQ of 120 (UK norms); 122 (U.S. norms) suggesting Kirk has a mean IQ in the mid 120s in my humble opinion.
Thus the biodemographic prediction is confirmed by actual psychometric performance on a highly g loaded mental task (debating in real time on diverse subject matters) that can be objectively graded by what Jensen would call a clear standard of proficiency.
The most interesting debate was with the medical student who challenged Kirk’s opposition to abortion. I think Kirk was right to argue that life begins at conception because that’s the only point in development where there is a discrete non-arbitrary separation between life and non-life. Everything after that is just a perfect continuum that can’t be objectively subdivided. However his weak point was failing to adequately explain what it is about life that makes it morally sacrosanct. Nonetheless despite the cheers for his opponent, I think Kirk convinced a lot of rational conservatives that abortion should be opposed, not for moral reasons where his arguments made no sense, but to avoid demographic replacement.
Commenter Teffec asked me to estimate the IQ of Charlie Kirk and I think the best way to do that is to first estimate the IQ of elite talk show hosts in general, since Kirk clearly falls into that category with several million subscribers on YouTube, even before his tragic death.
When Time magazine asked Meredith Vieira her SAT scores she replied 1300s which would equate to an IQ of around 136 in her day (U.S. norms). Although once part of a very famous talk show (The View), I don’t consider her an elite talk show host since The View is an ensemble, so only one or two of them needs to be talented for the show to work.
Joe Rogan posted a certificate on Instagram from the Brain Metrics Initiative showing his IQ was 127 which sounds about right. Smart enough to be the most successful podcaster in the World with a lot of thoughtful political views, but still dumb enough to be short and stumpy and somewhat gullible (or at least pretends to be since pseudoscience sells). Not sure how accurate the BMI test is but I seem to recall Rogan saying he never took the SAT which means he would have scored below an equivalent of IQ 135 had he taken it because Richard Herrnstein and Charles Murray along with Ron Hoeflin have argued that everyone who would have scored extremely high on the SAT, actually took the test (probably less true in recent years thanks to the ACT).
Meghan Kelly told Howard Stern she scored at the 85th percentile on the SAT which equates to about the 95th percentile if all Americans took it which equates to an IQ of about 125 which sounds about right. She’s bright, very bright, but not brilliant. It’s unclear whether she meant the verbal, math, or combined score.
Howard Stern claims to have scored 900 on the SAT which in his day equated to an IQ of about 110. I used to add bonus points since he attended a ghetto school but the famous Coleman report in the 1960s found school quality contributed very little to standardized test scores at least in Howard’s day. Many people would consider this a huge underestimate given Stern’s wit, humour, fluency and incisive interviewing style. It’s also way less than you’d expect from his most salient biodemographics. On the other hand, the hosts of Quite Frankly: A Howard Stern Podcast are constantly calling him a dumbass with an IQ of 79 so subjective impressions at both extremes tend to cancel out.
The average of all four of these people is 125 (U.S. norms) with a standard deviation (adjusted for degrees of freedom) of only 10.79, compared to the general U.S. population set at 100 (SD = 15). The ratio of their squared SD divided by America’s squared SD implies only 52% of the variation in IQ remains once you control for occupation, leaving 48% explained. The square root of 48% implies a potent 0.7 correlation between occupation and IQ, thus corroborating much previous research (Jensen 1998).
Of course self-reported data, especially from famous people should be treated with great caution as there is selection bias in who chooses to disclose their test scores, which test scores they choose to disclose, not to mention outright lying. On the other hand, using Jonathan Wai’s method of counting the percentage who attended elite schools, I had previously estimated that in 1994, the 11 top talk show hosts in U.S. television syndication had a mean IQ of 133 (U.S. norms) so there’s no reason to think ascertainment bias or selective disclosure inflated the average IQ in my admittedly tiny sample.
The data in the below chart comes from popular Youtuber Mason the Man. In a video called “The Formation of the World’s Modern Races” he identifies 17 races and estimates the age and population of each one. I don’t agree with his age estimates because in some cases they seem to refer to the “splitting off dates” and other cases they refer to how long they have looked the way they do now and these are two distinct concepts.
His population estimates sound more plausible, especially since when I added them all up, the total was 7.055 billion which was very close to the World population eight years ago ( 7.615 billion ) which is when the video was published. To make up for the shortfall I added an 18th category to the below chart (mixed race/other) though I suspect a lot of the non-whites races already include a lot of us hybrids and some of his races actually are hybrids (Pardo people of Brazil), actually all of them probably are if you go back far enough.
It’s my favorite time of the year. The leaves are changing color, hot chocolate is being served, horror films are streaming , and most important of all, Forbes just released their annual ranking of the 400 richest Americans.
The bad news? Just when blacks other than Oprah started becoming billionaires, billionaire status is no longer enough to be super rich. Sadly, as soon as a black achieve something, it’s no longer prestigious, kind of like when blacks move into a neighborhood, the house values decline because people move out. When a black became President, suddenly it was possible for orange game show hosts to do so too, and now that some blacks are finally billionaires, billionaire status is no longer that rich.
In fact, over a decade after Oprah became the first multi-billionaire black in American history, even several billion dollars is no longer enough to be super rich. This year you need an astonishing $3.8 billion just to rank dead last behind hundreds of much richer people.
One famous heiress from a very wealthy old money family wanted Forbes not to include her on their list so she could stay anonymous, only to be told “don’t worry you don’t qualify”. The woman went absolutely ballistic calling the Forbes reporter a “F@CKING C@NT! F@CKING WH@RE!!!!”.
Although some Old Money prefer their wealth to be downplayed, they want this to be their choice. Old money is very concerned that they’ve been kicked out of the rich club by a new generation of snot nosed computer geeks and investment bankers who they simply don’t have the IQ to compete with.
Despite being 13% of America, only 1% of the Forbes 400 (4 people) is black this year and that actually represents an all time high. For most of the list’s history, there were either no blacks or just one (usually Oprah) however even this all time high of 1% is misleading because half of them are arguably mulatto or even what racists used to call quadroon.
For centuries it’s been known that brain size is at least slightly linked to IQ which from a Darwinian perspective, should be at least slightly linked to wealth. Thus it’s absolutely fascinating that for decades, the richest black in America donned freakishly huge hats.
More recently, it’s been suggested that Ashkenazi ancestry is linked to high IQ (via genetic diseases like Tay Sachs) and so it’s equally fascinating that the man who has replaced Oprah as America’s richest “black” has a Jewish father and is probably the most intelligent black billionaire since Oprah herself. Although his politics are unevolved, he boasts a bachelor of Arts/Science from Haverford College and Doctor of Jurisprudence from Stanford University.
With a net worth of 11 figures, he’s far richer than Oprah ever was, but is he adaptable enough to hold the title of richest black in America for decades like she did, or is he just another flash in the pan, who will be pushed into the dustbin of black history by the next generation of new money? .
If AI just regurgitates and recombines human knowledge, does that make it dumber than us? Not necessarily because we humans also just regurgitates and recombine what we’ve learned from other humans. Has even the most brilliant human ever discovered anything that is not in some way a synthesis of what other humans or pre-humans already knew?
But then how will we know when AI is smarter than us? When it beats us on IQ tests? It already does in the skills it’s been trained on like verbal and mathematical reasoning but scores poorly in many other tasks. But then even the dumbest of humans would score high on an IQ test if endlessly trained. So how will we know when AI is truly smarter than us, and it’s not just an artifact of teaching to the test or mirroring our own intelligence back at us? This very brief video will tell you.
Pumpkin Person 