JFK, Nixon & the Otis Gamma IQ test

As a teen I believed the World was run by super-geniuses, but after years of emailing super high IQ people and doing more of my own research, I gradually realized that IQ is to political and economic power as height is to physical power. Just as the World’s most physically powerful man (i.e. heavy weight boxing champ) is typically about 2 standard deviations taller (about 6’3″) than the average American man (about 5’10”), the World’s most powerful man (U.S. President) is typically about 2 standard deviations smarter (IQ 130) than the average American man (about IQ 100). The latter assertion was mostly based on averaging the IQs of JFK (119) and Richard Nixon (IQ 143) both of whom had taken the Otis allowing an apples to apples comparison.

But commenter Ridin with Biden informed me of the following Quora comment:

Kennedy took the original version of the Otis Test, which was in vogue during the first half of the 20th century. The median on that test was 100, and the standard deviation was 10 points. Sometime later in the 20th century, the proprietors of the original test replaced it with a new version, called the Otis-Lennon School Ability Test (OLSAT). But scores between the two tests are not directly comparable, because the OLSAT has a standard deviation of 16.

As for the Wechsler – which is probably the most widely known and respected test – the median is 100, and the standard deviation is 15.

Hence, on that old, original Otis, a score of 119 corresponds with 96th or 97th percentile of all test takers. A Wechsler score of 128 is the 96th or 97th percentile of that test also.

But if the smaller SD upgrades Kennedy to 128 (actually 129), then it should also upgrade Nixon to a freakishly high 165, suggesting the average IQ of U.S. Presidents might be 147! Perhaps the World is run by super geniuses after all!

I decided to do a bit of research.

I found this study showing the Otis did indeed have a standard deviation around 10, at least in children:


But what about teenagers, since this was the age that Nixon and JFK were when tested. For this I turned to the Otis Gamma which was used with older kids and adults.

It should be noted that unlike most tests of that era, which calculated IQ as the ratio of mental age to chronological age, the Otis (or at least the Otis Gamma) assigned IQs based on how far you deviated from the mean of your age group, so if your raw score on the test was 43 points above the mean for your age (as Nixon’s apparently was) you were assigned an IQ of 143. This should not be confused with the deviation IQ where the standard deviation is held constant from age to age.

The test author noted that Gamma IQs found by this method tend to be “somewhat less variable than ordinary IQ’s”; “somewhat closer to 100”. However given that the SD of old Standford Binet age ratio IQs was 16.4, “somewhat closer to 100” needn’t imply less than 15.

I then found a 1971 study where 45 men from the University of Oklahoma part-time employment sample took the Otis Quick scoring Gamma test form C and the 40 minute Raven Advanced Progressive Matrices:

The raw Otis scores ranged from 43 to 76 with a mean of 61.6 and an SD of 9.2. Since the mean score for 18+ year-olds was 42, the assigned IQs ranged from 101 to 134 with a mean of 120. Meanwhile the Raven scores ranged from 14 to 35 with a mean of 24.9 and a standard deviation of 4.6. Using 1962 norms, a score of 14 = 75 percentile and 24 = 95 percentile (a difference of one standard deviation on a normal curve).

Assuming the lowest Otis Gamma IQ (101) of this group was also one standard deviation below the average score (120), then an SD of 19 is implied! Of course extrapolating from just two data points is unreliable and further complicated by the Flynn effect (Raven norms were a decade old and who knows how old the Gamma norms are) and the likely non-interval nature of raw test scores.

Nonetheless, it seems unlikely that the Otis Gamma had an SD as low as 10 (at least in adults).

Update Nov 10, 2020

A quick way of converting Otis raw scores to IQ at different ages:


Update Nov 12, 2020

The following source claims the Otis Gamma had an SD of 12.

Source: Mason, E. P., Adams, H. L., & Blood, D. F. (1966). Personality characteristics of gifted college freshmen. Psychology in the Schools, 3(4), 360–365.

If so, JFK’s 119 IQ would become 124 and Nixon’s 143 IQ would become 154 on the 15 sigma scale.

There’s also this:

Source: Traxler, A. E. (1934). Reliability, constancy and validity of the Otis IQ. Journal of Applied Psychology, 18(2), 241–251. doi:10.1037/h0075882 

The 1937 ediction of the Stanford Binet had a mean IQ of 101.8 with an SD of 16.4. Assuming the 1916 edition had the same distribution and was created four years earlier than the Otis (which appeared in 1920), we’d expect it to have a mean of 102.36 at the time of the Otis norming (The Stanford-Binet Flynn effect from 1916 to 1937 was only 1.4 points per decade). Thus:

Otis IQ 111 = +1.01 SD Binet

Otis IQ 108.3 = +0.95 SD Binet

Otis IQ 114.8 = +1.15 SD Binet

Otis IQ 111.4 = +1.05 SD Binet

The line of best fit suggests 9.91 Otis IQ points equates to 1 Binet SD, suggesting the Otis had an SD of 9.91. This suggests JFK’s IQ would be 129 on the 15 sigma scale and Nixon’s would be a superlative 165. But perhaps at this high extreme, Otis scores were not normally distributed and like a Binet IQ of 171 (recall the Binet mean was a bit above 100 and the SD was 16.4) which had about a one in 5000 level rarity, thus equating to a normalized IQ of 153 (sigma 15). These IQs might also be slightly inflated by the Flynn effect ao JFK was perhaps 126 and Nixon was perhaps 152, suggesting U.S. presidents average 139.

Gene editing IQ

Commenter Ganzir writes:

Step 1: Locate the single nucleotide polymorphisms associated with higher IQ; not necessarily all or even most of them, just a moderate number would be enough

Step 2: Use genetic editing to change a T here, a G there, and A or a C here… preferably on a genome from somebody who had a high IQ already for best results; this would probably create an individual with a higher genetic “compound score” for IQ than the randomness of natural breeding ever has

Step 3: ???

Step 4: Profit (after they earn all the million-dollar prizes from the Clay Mathematics Institute)

I guarantee you the Chinese government is already walking this path, and there’s still time to change the road we’re not on. I advise not to waste it.

The problem with this is there are probably about 10,000 genetic variants associated with IQ so you’d would have to edit about 1000 of them just to raise IQ 10 points. And who knows how many other traits you might accidentally damage while raising IQ. Not to mention, many of these variants will be expressed in the growing brain so if you alter them in adulthood after the cranium locks tight, the brain might get too big for it’s skull, causing death.

But once they do figure out how to edit an adult’s DNA to increase his IQ by say 30 points, it would be fascinating to give said adult a battery of cognitive tests.

For years commenter “Mug of Pee” has belittled psychologists for believing intelligence tests can be classified as crystallized (acquired knowledge) vs fluid (novel problem solving). As Arthur Jensen noted, a typical fluid item might be Grass is to cattle as bread is to? while a typical crystallized item might be Pupil is to teacher as Aristotle is to?. In the first question, almost everyone has enough general knowledge to solve it so the variation in scores comes from making an inference within the test room. In the latter, almost everyone has enough reasoning in the test room to solve it, but the variation in scores comes from a lifetime of different rates of learning through inference so by the time you get to the test room, you know who Socrates was and how he related to other historical figures.

So fluid tests are thought to observe intelligence directly, while crystallized tests indirectly measure it. The indirect measures can sometimes be more accurate but they are lagging indicators which should be most noticeable in those who experience an organic change in cognition, at least in theory.

So if we could increase a person’s genetic IQ and then test them before they had time to update their neural hardware with the software of life experience, I predict they would perform much better on the so-called fluid tests than on the so-called crystallized ones. Of course I would expect some improvement on all tests because even experience based tests require the fluid retrieval of information and the ability to organize it.

About a decade after the gene editing I would expect the gap between fluid and crystallized to more or less close as the new and improved brain acquires knowledge and vocabulary at a faster rate.

How accurately will DNA predict height & IQ?

Jocelyn Kaiser of sciencemag.org writes:

For height, DNA is largely destiny. Studies of identical and fraternal twins suggest up to 80% of variation in height is genetic. But the genes responsible have largely eluded researchers. Now, by amassing genome data for 4 million people—the largest such study ever—geneticists have accounted for a major share of this “missing heritability,” at least for people of European ancestry. In this group, they’ve identified nearly 10,000 DNA markers that appear to fully explain the influence of common genetic variants over height…

…By 2018, Visscher’s team and other members of a global consortium called GIANT had pooled DNA data for 700,000 people and found 3300 common markers that explained 25% of the variation in height. Now, by looking across DNA from 201 GWA studies with 4.1 million participants, GIANT has brought the total to roughly 9900 common markers, accounting for 40% of the variation. Other markers located nearby and likely inherited together account for another 10% of height variability.

That’s still short of the 80% predicted by twin studies. But last year, Visscher’s group drew on whole-genome sequencing data of a smaller number of people to demonstrate that rare variants—those carried by fewer than one in 100 people—should explain another 30% of height’s variation. (The result was released in a March 2019 preprint that the team is revising.)

Some geneticists say they aren’t surprised that heritability gaps can be filled once enough people had their DNA scanned. “It was expected,” says Aravinda Chakravarti of New York University. The problem remains that few of the height-linked DNA markers have been tied to specific genes that clearly alter the trait. “It’s mostly all still ‘missing’ in a biological sense,” says David Goldstein of Columbia University.

This is exciting news. Even though they still can’t say what genetic markers cause height, they will soon be able to predict it with incredible accuracy from just your DNA (at least for modern Westerners). If genomic markers explain 80% of the variation, the correlation between height and DNA would be 0.89! (the square root of 80%). In practical terms that means you could guess someone’s height just from DNA alone and be within 2.6 inches 95% of the time (at least for whites living in the West and eventually for all major populations).

Can we expect similar results for IQ? In the Minnesota study of twins reared apart, the square root of IQ’s heritability was about 92% as large as height’s. So if polygenic scores will one day correlate 0.89 with height, then perhaps a correlation of 0.89(0.92) = 0.82 can be expected for IQ (assuming a mental measure like IQ is analogous to a physical one like height).

This is about as high as the very best IQ tests correlate with each other, and implies they will be able to guess someone’s IQ based on DNA alone, and 95% of the time, those guesses will fall within 17 points of the correct IQ. In theory the precision could be increased if they predicted one’s composite IQ score on multiple high quality tests administered across the life span (thus cancelling out measurement error)

Of course none of this proves DNA causes most of the IQ variation but there is enormous practical value is predicting IQ, regardless of causation, even if said predictions are largely limited to specific nations and generations. Though if enough truly causal variants can be found, they will predict one’s cognitive rank in any society.


I’ll be live blogging the election. Check this page throughout the day & night for updates.

Update 1:27 pm Eastern: Back in 2016, the last poll I looked at before the election was Rasmussen which showed Hillary up by a couple percentage points nationally (which was technically accurate since she won the popular vote). I figured Biden was doing way better against Trump than Hillary did, so to my surprise, Biden is leading only 1 point: 48% vs 47%.

Update 2:22 pm Eastern: The Washington Post reports:

Biden leads Trump by 10 percentage points nationally, 52 percent to 42 percent, according to an average of national polls since Oct. 26. Biden’s margin in the battleground states of Michigan and Wisconsin is nine points. It’s five points in Pennsylvania, five in North Carolina, four in Arizona and three in Florida.

Update 8:43 pm Eastern With 49% of the vote in, Biden is beating Trump in Ohio 55% to 44%. No Republican in modern times has ever won without Ohio.

Update 9:06 pm Eastern Biden beating Trump 73 to 48 electoral votes. 270 needed to win.

Update 9:27 pm Eastern Both men tied at 49.3% in Texas

Update 9:43 pm Eastern Looks like Biden’s lead in Ohio has shrunk down to 49.5% to 49.2%.

Update 10:04 pm Eastern Trump now leading in Ohio: 50.8% vs 47.9%

Update 10:16 pm Eastern Trump now leading in 3 of 4 battleground states

Update 10:25 pm Eastern Trump now leading in Michigan, Wisconsin, Pennsylvania, North Carolina, and Florida. Biden leading Arizona. This is disturbing because the polls predicted Biden was ahead in all of these places.

Nov 4, 2020

Update 2:18 pm Eastern: All Biden has to do now to hit 270 is maintain his lead in Michigan, Arizona and Nevada. Pumpkin Person is now officially calling the race for Joe Biden. He will be the next President of the United States.

Slasher IQ & Behavioral Modernity

Scientists use the term Behavioral Modernity to describe a suite of sophisticated traits thought to have evolved around 50,000 years ago, marking the start of the Upper Paleolithic. For example Richard Klein noted that before 50,000 years ago, virtually everyone made artifacts out of stone, while after 50,000 years ago, folks suddenly started using bone, ivory, antler or shell. He cited this fact, among many others, as evidence of a genetic increase in intelligence that occurred in our species.

If using diverse materials to make artifacts is evidence of more intellect among cavemen, perhaps using diverse weapons to kill people is evidence of more intellect among slashers. In honor of Halloween, I decided to see how some of our favorite fictional slasher villains would stack up.


In Friday the 13th part II (1981) (the first film that features Jason as a slasher), Jason kills 8 people with 6 weapons (ice pick, barbed wire, hammer claw, machete, spear, and butcher knife). That’s a ratio of 6/8 or 0.75.

Michael Myers

In the original Halloween (1978), adult Myers kills 4 people (and two dogs) with at least 3 weapons (knife & phone chord): Score 2/4 = 0.5


She kills 8 people with 5 weapons (hunting knife, arrow, bow and arrow, machete, and axe). Score 5/8 = 0.63


In Silent Night Deadly Night (1983), he kills 8 people with 6 weapons (Christmas lights, knife, hammer, bow and arrow, deer horns, axe). Score 6/8 = 0.75.


In the original Texas Chainsaw Massacre (1974) he kills 4 people with 3 weapons (hammer, hook and saw). Score 3/4 = 0.75

Distribution of scores

Mean Behavioral Modernity Quotient = 0.68 (Standard Deviation = 0.1)

On a scale where British people average IQ 100 (standard deviation = 15), the criminally insane probably average about 76 (SD = 15?).

By equating the Behavioral Modernity mean and SD of the slashers (0.68 and 0.1) to the criminally insane IQ mean and SD (76 and 15), we get the following rough measures of non-verbal IQ:

Jason = IQ 87

Michael Myers = IQ 49

Pam = IQ 69

Billy = IQ 87

Leatherface = IQ 87

These numbers don’t seem to correlate at all with the intelligence of these characters. But a low correlation is to be expected given the small and fictional nature of the sample. More research on large data-sets of real life serial killers might give better results.

Hereditary (2018)

Pumpkin Person rating: 7 out of 10

For years commenter “Race Realist” has been telling me to watch this movie (because I’m a horror fan who is interested in genetics?) but I avoided it because when it comes to horror, I prefer the semi-realistic (i.e. slasher films) to the supernatural (i.e. ghost stories) because the former at least have some kind of logic to then, while in the latter, anything goes (people start flying or shape-shifting, pillows turn into people).

But in the spirit of Halloween I decided to step out my comfort zone and give this film a chance.

The film tells the story of Annie Graham (played by Toni Collette), her husband Steve ( Gabriel Byrne ), and their two teenaged kids, Peter (Alex Wolff ) and Charlie ( Milly Shapiro). For a film called Hereditary, there’s not a lot of physical resemblance between the swarthy Wolf and the pale actors who play his parents.

The Grahams are an upper middle class family with a creepy but beautiful home in the Woods. The film starts with the death of Annie’s mother (who suffered from mental illness) and her fear that her daughter Charlie will do so too. She encourages her socially well-adjusted son Peter to take Annie to a party and something utterly horrific happens. It happens so quickly that I actually had to rewind the film several scenes later to see exactly what happened because I had no idea I had missed something that important.

The film is well-written, well-acted and well-directed (with lots of use distant camera shots) but it’s not fun, enjoyable or entertaining. But if you’re looking for an atmospheric and effective horror film to watch on Netflix this Halloween season, it might be worth a look. If you’re looking for a horror film about Behavioral Genetics, don’t let the title fool you.

What effect does race have on income?

Commenter “destructure” wrote on Lion of the Blogoshphere’s blog:

 blacks with an IQ of 100 actually earn the same as others with an IQ of 100. That’s fine. However, blacks with an IQ of 120 earn more on average than others with an IQ of 120. And I’m not sure why upper middle class blacks should be getting affirmative action when they’re already earning more money than others.

Lion responded:

Totally makes sense because a black with an IQ of 120 will get accepted to a better college, get affirmative-action hired into a better career, than a white with the same IQ. But what’s the source for that factoid?

On the other hand, if high IQ blacks overperform their white counterparts economically, we’d expect more black billionaires. At yet despite being 13% of America, blacks have never been even 1% of the Forbes 400 richest Americans list and are sometimes only 0.25%. By contrast Jews are arguably 36% of the Forbes 400 despite being only 2% of America (though exact stats depend on how Jews are counted).

I decided to use my celebrity status to reach out to the World’s most influential social scientist, Charles Murray, since his book Human Diversity found that even controlling for IQ, white Americans are more prosperous than black Americans, though less so than Asian Americans (at least among Generation Y thirty somethings). With respect to high IQ blacks in particular, Murray replied:

But if high IQ blacks benefit from affirmative action, why doesn’t that translate into economic benefits? One possibility is that affirmative action devalues black credentials. For example Supreme Court Justice Clarence Thomas was overlooked by dozens of law firms despite having one of the most prestigious law degrees in the country. Another possibility is that whatever edge blacks get from affirmative action is negated by racism.

On the other hand, if America is still racist, why do Asian Americans earn 157% as much as their IQ matched white counterparts according to Murray’s book? Perhaps because they work much harder than whites do while experiencing much less racism than blacks do (at least prior to the coronavirus pandemic).


It’s also worth noting that whatever advantage Asian Americans have in regular income has not yet showed up among the Forbes 400 suggesting that hard work only goes so far.

Favorite scene in the Halloween franchise

With Halloween only 2 days away , I thought I’d post my favorite scene in the entire franchise. It’s from Halloween II (1981) and features a guy bumping into Michael Myers on the street while listening to news coverage about the Myers massacre on his radio.

What I love about this scene is that the whole town obsessing over Myer’s killing spree, he’s all over the news, everyone’s talking about him, police are racing through the streets to find him, and Myers is just casually walking through the heart of downtown and no one notices. Talk about hiding in plain sight.

I also love the look of the mask from behind and how the hair on it has an almost reddish hue in the city street lights. The Myers mask has never looked better than in Halloween II. It’s the same bleached William Shatner mask that they used in the original, but something about the way it fits like a glove on this actor’s head makes it look so beautifully demented and dare I say, autistic.

Hardcore cross-cultural IQ testing

For years I have dreamed of making the most culture-fair IQ test possible and then venturing into the Australian Outback, the African jungles, and the Canadian arctic and testing the natives. Sadly, the coronavirus has derailed those plans from happening, but I was excited to read about similar research published by J.W. Berry in 1966 and 1971. Berry tested five different “races” (Scotts, Eskimos, New Guinean aboriginals, Australian aboriginals, the Temne), each reared in two types of environments (Traditional vs Transitional) on at least three different tests (Khos Block Design, Embedded Figures, and Ravens Matrices).

Ethnic group
Traditional environment
Khos Block Design
Traditional environment
 Embedded Figures
Traditional environment
Transitional environment
Khos Block Design
Transitional environment
Embedded Figures
Transitional environment Matrices
 IQ 97  IQ 92 IQ 92 IQ 97 IQ 97 IQ 97
 Eskimos IQ 91  IQ 92 IQ 82 IQ 97 IQ 96 IQ 90
 New Guinea (Indigene) IQ 59 IQ 41 IQ 39 IQ 84 IQ 87 IQ 81
 Australian aboriginal IQ 73 IQ 81 IQ 71 IQ 78 IQ 83 IQ 75
 Temne IQ 57 IQ 41 IQ 33 IQ 62 IQ 49 IQ 36
IQ were calculated by converting raw scores into Z scores using the mean and standard deviation of the transitional Scottish (Edinburgh) sample (table 7 of Berry, 1966)) and then multiplying by 15 (the standard deviation for IQ) and adding 97 (the estimated IQ of Scotland). The raw scores for the Scottish, Eskimo and Temne samples were also found in table 7 of Berry 1966 and and the raw scores for the New Guinean Indigene and Australian aboriginal samples was found in table 2 of Berry 1971; scores of men and women were averaged. The IQ gap between the highest and lowest scoring samples might be exaggerated by the fact that the distribution has not been normalized. On the other hand, the raw scores from which IQ were calculated were based on ages 10 to 70 combined into one sample which tends to inflate the standard deviation and thus minimize IQ gaps.

The absurdly low IQ of some groups on some tests highlights the challenge of creating “culture fair” tests that give credible results. While the low scores on Block Design might be explained by a combination of genetics and severe malnutrition (Sierra Leon men are 2.82 standard deviations shorter than their Black American counterparts), the sub-50 IQ means on Embedded figures and sub-40 IQ means on Matrices leaves little doubt that at least those two tests are culturally biased.