The Kalash are a fascinating population because they look just like whites, yet are indigenous to South Asia. Genetically, they are more distant from Whites than Whites are from the blackest skinned South Asian which is why I’ve long argued that genetics is a poor way to define race.
Human populations show subtle allele-frequency differences that lead to geographical structure, and available methods thus allow individuals to be clustered according to genetic information into groups that correspond to geographical regions. In an early worldwide survey of this kind, division into five clusters unsurprisingly identified (1) Africans, (2) a widespread group including Europeans, Middle Easterners, and South Asians, (3) East Asians, (4) Oceanians, and (5) Native Americans. However, division into six groups led to a more surprising finding: the sixth group consisted of a single population, the Kalash
So isn’t that interesting that when you divide the human genepool into five clusters, you get traditional races: 1) Negroids, 2) Caucasoids, 3) Mongoloids, 4) Australoids, and 5) Americoids. But when you divide us into 6 clusters, the Kalash emerge as their own distinct macro-race. This shows that the split between Kalash and conventional Caucasoids is about as deep and as ancient as the paleolithic splits between major races like Native Americans and East Asians and predates civilization, agriculture and even the Holocene.
Since the split from other South Asian populations, the Kalash have maintained a low long-term effective population size (2,319–2,603) and experienced no detectable gene flow from their geographic neighbors in Pakistan or from other extant Eurasian populations. The mean time of divergence between the Kalash and other populations currently residing in this region was estimated to be 11,800 (95% confidence interval = 10,600−12,600) years ago, and thus they represent present-day descendants of some of the earliest migrants into the Indian sub-continent from West Asia.
If the Kalash diverged from the ancestors of whites 12,000 years ago, yet look just like whites, then either a white looking phenotype evolved twice independently, or much more likely, the white race is at least 12,000 years old.
Whites should be very proud to be 12,000 years old! Young enough to imply evolutionary progress (if you believe in such) but old enough to have been selected by nature (before agriculture and civilization).
It would be nice if some of the white nationalist types who are so concerned about preserving the white race, would put some of that energy into preserving the Kalash who are far more at risk of extinction and represent the last representatives we have of original whiteness.
With a mean IQ of 100, whites are one of the highest IQ groups on the planet (behind only Ashkenazi Jews and East Asians). But were their genetic IQs 100 from inception, or did they only become 100 after agriculture and civilization. We know for example that Native Americans and Arctic people score lower than their East Asian cousins, suggesting that the neolithic transition might have boosted IQ.
Thus I would predict that the Kalash (even if raised from birth in middle class Western society) would score at least 0.5 SD below conventional Whites.
It’s also interesting that the Kalash are some of the earliest migrants into South Asia. Is it possible that South Asians evolved from whites?
A reader stated provided a screenshot of his performance on humanbenchmark.com.
The reader states:
… humanbenchmark.com, that website where you test your reaction speed, has a wide selection of other psychometric tests, I’d guess a composite score of all the tests would probably have a decently high g-loading. I just want some background info on these tests, if there is any.
As discussed in previous articles in this series, some of the tests (sequence memory, number memory) have their roots in conventional psychometric tests. Tests of reaction time date back to the 19th century work of Francis Galton who believed that basic neurological speed predicted intelligence. Unfortunately Galton’s research was derailed by a lack of reliability (he only used a one trial measure of reaction time) range restriction (his samples tended to be elite) and improper measures of intelligence with which to relate reaction time (he compared it with school grades since IQ tests had not yet been invented). As a result, he detected virtually no relationship between reaction time and intellect.
Nearly a century later Arthur Jensen would revisit Galton’s work, correcting for these problems. He found that when you aggregated many different kinds of reaction time (simple, complex, etc) measured both by speed and consistency (faster and less variable RTs imply higher intelligence) over many different trials, and compared with measures of IQ (not grades) and corrected for range-restriction, the results correlated a potent 0.7 with intelligence.
Unfortunately, the human benchmark test only uses simple reaction time (which is much less g loaded than complex RT), only one type of simple reaction time (an aggregate of several types is more g loaded) and only measures speed (variability is much more g loaded) and does not provide a composite score weighted to maximize g loading. As a result, on the whole the human benchmark tests seem inferior to the game THINKFAST which a bunch of us played circa 2000. So accurate was THINKFAST that the Prometheus society considered using it as an entrance requirement, with one internal study finding that one’s physiological limit on THINKFAST correlated a potent 0.7 with SAT scores in one small sample of academically homogenous people. Having people practice until hitting their physiological limit was a great way to neutralize practice effects because everyone must practice until their progress plateaus.
Sadly, this innovative research petered out when people worried that Thinkfast might give different results depending on the computer. People fantasized about Thinkfast being on a standardized handheld device so scores could be comparable, but in those days, few people imagined we’d one day all have iphones and ipads.
The reader continues:
I’ve also attached a screenshot of all my average scores, though I’ll note that some scores are inflated since I’ve done all the tests many times and I often don’t bother finishing the test if I do bad. The strange thing about these scores is that by more conventional measures both my verbal IQ and working memory are pretty average, yet I’m able to score above the 99.9 percentile on 2 of these tests. I think this points to the fact that memory is an ability that is much broader than most IQ models would suggest. Like the verbal memory test in particular, I seem to be using a very different part of my brain compared to more typical tests like digit span. I’d also wager that most of the variation in working memory can be explained by chunking/processing abilities rather than raw storage capacity. Also, what does the strength of the practice effect really say about a test? None of these tests really have a pattern or trick to them, yet for some of them my score has improved a lot from the first time I did them.
This is an extremely important question. In complex cognitive tasks like chess or conventional IQ tests, practice improves performance because we learn strategies, but on elementary cognitive tasks like Human Benchmark and Thinkfast, fewer strategies are possible so one wonders if there’s an increase in raw brain power.
The analogy I make is heigt vs muscle. If I repeatedly had my height measured, I might score a bit higher with practice. Not because I was genuinely getting taller, but because I was learning subtle tricks like how to stand straighter. By contrast if I had my strength measured everyday, I’d show more increase, but this increase would not simply be because I acquired tricks to do better (how I position the barbells in my hands) but because a genuine increase in strength.
So is intelligence more analogous to height or physical strength (the latter being far more malleable)? Is the practice induced increase in Human Benchmark tests an acquired strategy (even a subconscious one) or a real improvement, and how do we even operationalize the difference?
If practicing elementary cognitive tasks really did improve intelligence we’d expect brain-training games to improve IQ, but apparently they do not. Jordan Peterson explains that the problem is that cognitive practice in one domain does not translate to other ones.
On the other hand, why should anyone expect brain training to transcend domains? When a weight lifter does bicep curls, he doesn’t expect it to make his legs any stronger, so why should someone practicing visual memory expect to see an increase in verbal memory, let alone overall IQ?
But how can we know if we’ve even improved a specific part of intelligence rather than just become more test savvy? We know that weight lifting has improved our strength, and not just our technique, because we can see our muscles getting bigger, so perhaps cognitive training games might make certain brain parts bigger.
The groundbreaking London Taxi Cab study, published in 2000, used MRI technology to compare the brains of experienced taxi cab drivers and bus drivers who drive the city streets of London every day. In contrast to bus drivers, whose driving routes are well-established and unchanging, London taxi drivers undergo extensive training to learn how to navigate to thousands of places within the city. This makes them an ideal group to use to study the effects of spatial experience on brain structure.
The study focused on the hippocampus, which plays a role in facilitating spatial memory in the form of navigation. The MRI revealed that the posterior hippocampi of the taxi drivers were much larger than that of the bus drivers (who served as the control subjects). Even more exciting was that the size of the hippocampus directly correlated with the length of time that someone was a taxi driver–the longer someone drove a taxi, the larger their hippocampus.
The London Taxi Cab Study provides a compelling example of the brain’s neuroplasticity, or ability to reorganize and transform itself as it is exposed to learning and new experiences. Having to constantly learn new routes in the city forced the taxi cab drivers’ brains to create new neural pathways “in response to the need to store an increasingly detailed spatial representation.” These pathways permanently changed the structure and size of the brain, an amazing example of the living brain at work.
Assuming the brains of the taxi drivers actually changed (as opposed to the sample changing because less spatially gifted drivers left the job) it might be possible to increase specific parts of intelligence, but since there are so many different parts, it’s perhaps impossible to ever increase overall intelligence (or overall brain size) by more than a trivial degree. We can improve our overall muscle mass because our muscles are outside or skeleton; by contrast our brains our inside our cranium so its growth is constrained. It could be that improving the size of one part of the brain requires a corresponding decrease in other parts, to avoid the overall brain from getting too big for its skull.
My research assistant 150 IQ Ganzir also weighed in on the reader’s questions, writing:
The first aspect of this score profile I noticed is the absence of any huge dips, the 10 on Number Memory notwithstanding, since a tiny change in raw score on that test can dramatically alter your percentile ranking. Given that all of this subject’s scores on the more IQ-like tests are well above average compared even to other HumanBenchmark users, who themselves are undoubtedly self-selected for superior proficiency on these types of tasks, we wouldn’t expect their reaction time to be particularly fast, but it is. Our subject appears to be a jack-of-all-trades, if you will, at these tasks. Simple reaction time has only a weak correlation of about -0.2 to -0.4 with IQ, according to Arthur Jensen on page 229 of The g factor. Note that the correlation is negative because a faster reaction speed implies a lower reaction time.
The commenter mentions: “I’ve also attached a screenshot of all my average scores, though I’ll note that some scores are inflated since I’ve done all the tests many times and I often don’t bother finishing the test if I do bad.” If true, this would indeed cause a statistical upward bias, but I have no idea how to even begin calculating the size of that. However, if the tests are reliable in the statistical sense, meaning they give similar scores with each administration, then the average score increase couldn’t be too large. But, then again, if the commenter was reaching nearly the same score every time, why would they restart on a bad run? High intra-test score variability might indicate executive functioning problems.
The commenter notes that their verbal IQ and working memory are “pretty average” on other tests, but their score on verbal memory here is so high relative to other HumanBenchmark users that the system just gives it 100th percentile without discriminating further. (I know that it can’t literally be 100th percentile, as I and several other people I know have achieved higher scores.) A possible contributing factor is that HumanBenchmark users may tend to have less than long attention spans, inhibiting performance on this test, on which reaching one’s potential may take quite a while, especially for higher scorers.
Our correspondent also writes: “Like the verbal memory test in particular, I seem to be using a very different part of my brain compared to more typical tests like digit span. I’d also wager that most of the variation in working memory can be explained by chunking/processing abilities rather than raw storage capacity.” Of course, I don’t think it’s possible to determine by introspection which part(s) of the brain you’re using on a given task, but I think I understand the subjective experience described here. As for chunking/processing abilities versus raw storage capacity, I’m not sure what’s implied here. The human brain could be described as a massively parallel computer, and it naturally processes things in chunks. If “chunking” refers to purposely learnt mnemonics, such as the mnemonic major system, then Goodhart’s Law applies here because learnt skills lose their g-loading.
The commenter thus wonders about the continued meaning of their scores: “Also, what does the strength of the practice effect really say about a test? None of these tests really have a pattern or trick to them, yet for some of them my score has improved a lot from the first time I did them.” Unfortunately, without studies of these tests specifically, we can’t know the extent to which Goodhart’s Law applies. Even analyses of seemingly similar tests from mainstream psychometrics wouldn’t be insufficient, since the HumanBenchmark versions are subtly but crucially different. All I can say is that only someone of uncommonly high cognitive capacity could produce this score profile regardless of how much time they spent practicing, and that, with no indication of how rare your scores are compared to the general population, greater precision is currently almost meaningless.
Scores on the “Chimp Test,” or at least the version on HumanBenchmark, are also almost meaningless because unlimited time is allowed to review the digits’ locations before answering, making it less a test of visual working memory and more a test of how long the testee is willing to stare at boxes. Also, most people will probably on average score higher on the HumanBenchmark “Number Memory” test than on the clinical version of the Digit Span test, since the former presents the digits simultaneously and allows a few seconds to mentally review them, whereas, in the latter, each digit is read only once with no opportunity for review.
Finally, the subject’s strong performances on Typing and Aim Trainer make me suspect a background in competitive computer gaming.
Obviously I can’t devote an article to every Human Benchmark test so I’m limiting myself to the best ones. One of the best is number memory.
Digit Span is measured by the largest number of digits a person can repeat without error on two consecutive trials after the digits have been presented at the rate of one digit per second, either aurally or visually. Recalling the digits in the order of presentation is termed forward digit span (FDS); recalling the digits in the reverse order of presentation is termed backward digit span (BDS). Digit Span is part of the Stanford Binet and of the Wechsler scales. Digit Span increases with age, from early childhood to maturity. In adults the average FDS is about 7; average BDS is about 5. I have found that Berkley students, whose average IQ is about 120; have an average FDS of between 8 and 9 digits.
The g Factor by Arthur Jensen, page 262
It should be noted that the Human Benchmark version of digit span does NOT include the Backwards version and shows all the digits at once for several seconds, not each one at a rate of one per second, and it only has one trial per level so there’s no room for error. For this reason I suggest taking your best score on your first two attempts.
So important is this test that it is one of the 10 subtests handpicked by David Wechsler himself for his original Wechsler scale, published in the 1930s.
Perhaps no test has been so widely used in scales of intelligence as that of Memory Span for Digits. It forms part of the original Binet Scale and all the revisions of it. It has been used for a long time by psychiatrists as a test of retentiveness and by psychologists in all sorts of psychological studies. Its popularity is based on the fact that it is easy to administer, easy to score, and specific as to the type of ability it measures. Nevertheless, as a test of general intelligence it is among the poorest. Memory span, whether for digits forward or backward, generally correlates poorly with other tests of intelligence. The ability involved contains little of g, and, as Spearman has shown, is more or less independent of this general factor.
The Measurement and Appraisal of ADULT INTELLIGENCE 5th edition, David Wechsler, 1958, page 70 to 71
On page 221 of The g Factor, Jensen notes that FDS and BDS have g loadings of about 0.30 and 0.60 respectively.
Wechsler goes on to explain that despite being a poor measure of intelligence overall, he included it in part because in his eyes, it’s a great measure of low intelligence: “Except in cases of special defects or organic disease, adults who cannot retain 5 digits forward and 3 backward will be found, in 9 cases out of 10, to be feeble-minded or mentally disturbed.”
The other reason he included it is he viewed it as an excellent measure of dementia.
I’m not convinced the test is better at low levels than at high levels. For example, Charles Krauthhammer towered with a spectacular of BDS of 12, and his genius is validated by the enormous influence he had over U.S. foreign policy.
In the below poll your level corresponds to the highest number of digits you correctly remembered on at least one of your first two attempts:
19% of my readers self-reported Human Benchmark sequence memory highest scores of level 21+ (after 10 attempts).
3% of my readers self-reported highest scores of 6 or less.
Evidence continues to accumulate showing that on a scale where Americans average IQ 100 (SD = 15), my global readership towers with an average IQ of 129 (SD = 19). Thus assuming a normal curve, the top 19% and bottom 3% should have IQs of 147+ and sub-98 respectively.
Assuming the sequence memory test is sufficiently g loaded, this implies level 21 = IQ 147 and level 6 = IQ 97.
Thus I would predict that a random sample of American youngish adults would average 6.84 (SD = 4.5).
Put simply:
IQ = 77 + (highest level obtained in first 10 tries)(3.33)
However one oddity about the self-reported data is that all of the people scoring 21+ score 24+. Nobody reported a score of 21 to 23. This suggest inaccuracy of self-reported data but it may also suggest that above level 21, the test starts measuring certain cognitive strategies and stops measuring g.
Humanbenchmark.com includes many tests, the first of which is called Sequence Memory. You can try the test here.
The way it works is nine squares appear on the screen and then some of them light up in sequential order. At level one, one square lights up. At level two, two squares light up. Level three, three squares light up. etc. Your job is to wait for the sequence to finish and then click on the squares in the same sequence they lit up. The test progresses in difficulty until you make a mistake.
It’s interesting to note that when the Wechsler intelligence scale for children (WISC) was first revised (WISC-R) Wechsler tried to add a subtest much like this. In the 1974 WISC-R manual he writes:
All twelve tests were administered to the entire standardization sample, as was also a thirteenth test entitled Visual Motor Memory. The latter test, requiring the child to copy a sequence tapped by the examiner, was intended as a supplementary Performance test and as a non-verbal visual motor analog of Digit Span. In spite of certain merits, the test was eliminated from the WISC-R because it posed problems of administration and failed to meet some of the rigorous statistical standards that were applied.
WISC-R manual page 8
I can see how it would pose administration problems because the examiner would need an excellent visual memory herself just to see if the child repeated the sequence correctly, especially if he did so quickly. But when the test is administered by computer like on humanbenchmark.com, that’s no longer an issue.
Nonetheless, the newest edition of the WISC (WISC-V) includes subtests like Picture Span and Spatial Span. In Picture Span, the child sees a bunch of pictures in correct order and then must remember their sequence when present in random order. This draws on similar cognitive skills as humanbenchmark’s sequence memory test and it correlates 0.61 with full-scale IQ (a strong proxy for g) among 16-year-olds. Oddly, the spatial span subtest (where a child must tap a bunch of blocks in the same sequence he sees the examiner do it) only correlates 0.43 (WISC-V technical manual, page 69), even though this includes a backwards spatial span component. When it comes to auditory sequencing, going from backward to forward doubles the g loading so I’d love to know why picture span is so much more g loaded than spatial span, despite having no backward component. Maybe there’s too much error because the examiners can’t keep up?
Humanbenchmark.com provides percentiles for scores on the sequence memory test, but there’s no context. Does the percentile reflect how many people we beat or how many attempts we beat (many by the same person) ? And how self-selected is the reference sample? It would be interesting to know how my readers (mean IQ 130) do on this test, how their scores respond to practice, and how this relates to their scores on established psychometric tests.
Are there any cases where an individual’s real-world achievement(s) align(s) with that individual’s stellar high-range test score? If so, what achievement(s), what score, and on what test?
The most obvious example is John H Sununu who exceeded the one in a million level is both Mega Test IQ and real-world power (state governor, White house Chief of Staff under a relatively good President: George H.W. Bush, who unlike his son, was competent enough not to invade Iraq). In 1988 the Washington Postreported:
President-elect George Bush’s new chief of staff, John H. Sununu, is a smart guy by all accounts. His doctarate from the Massachusetts Institute of Technology is one clue. His I.Q. is another. On a scale where “genius” is anything over 140, Sununu has been clocked at 176. That makes him one in a million.
Whether he is smart enough to massage the egos at the top of the administration and smart enough to befriend Jewish leaders and smart enough to outsmart his boss remains to be seen. But there is no question that he is book smart.
In 1985, Sununu picked up Omni magazine’s “mega-I.Q.” test while on a plane flight. At the time he was governor of New Hampshire and was busy with the state legislature, but when the session adjourned for the summer, Sununu got serious about the test. You might say it’s his idea of how to spend a summer vacation.
When he and more than 3,000 other Omni readers turned their tests in, Sununu had tied with two others for second place. He correctly answered 44 of 48 questions, where a score of 15 ranked the contestant as a genius with an I.Q. of 141. His score of 44 put Sununu’s I.Q. at 176.
“This test was one of the most enjoyable exercises I have gone through in some time,” Sununu wrote to the magazine, “a superbly stimulating diversion.”
It should be noted that the Mega Test annoyingly uses a standard deviation of 16, so an IQ of 176 is equivalent to 171 on the 15 sigma scale used by the Wechsler and recently even adopted by the Stanford Binet. It equates to the one in a million level. Since the Washington Post wrote that article Sununu’s score has been upgraded to the one in 3 million level.
Among the Omni reading test takers who reported their SAT scores, only two had 44+ Mega Test scores and only two had 1586+ self-reported SAT scores (not the same two), which at least in the 1980s, equated to the one in 700,000 level, so at the very least, I would put Sununu’s Mega score at +4.67 SD, especially since he achieved it on his first attempt and in the pre-internet era (before answers got leaked or became easier to research). Maybe post-Omni data allowed Ron Hoeflin to refine the score upward. Such a spectacular IQ gave Sununu the authority to belittle the intelligence of a man as intellectually respected as Obama.
Sununu is not the only U.S. governor with a reported test score equating to perhaps 170 (sigma 15). In 2006, Eliot Spitzer told the New York Times that he scored 1590 on the (pre-recentered) SAT and a perfect score on the (old) LSAT. In the Omni sample, I noticed Mega Test takers with self-reported old LSAT scores averaged 722 (SD = 30) on the LSAT and IQ 152 (SD 9) on the Mega Test. Assuming linear extrapolation, that would equate a perfect 800 LSAT with a Mega IQ of 175 (equivalent to 170 on the WAIS).
Sununu and Spitzer were not governor at the same time, and did not achieve their scores on the same test, but perhaps at any given time, about one of the 50 U.S. governors would score over +4.5 SD on any particular highly g loaded test (whether the Mega Test, college admission tests, or conventional IQ tests with sufficient ceiling). Similarly, one in 100 U.S. senators (Chuck Schumer) had a self-reported perfect or near-perfect SAT score, as does about one out of every 100 or so self-made members of the Forbes 400 richest Americans (Bill Gates, Paul Allen). If the top one or two percent of self-made real-world elites have IQs around +4.5 SD, it suggests the average (self-made) real-world elite has an IQ around +2 to +2.5 SD.
Sununu may not be the only Mega Test luminary who has achieved so spectacularly in the real-world. Circa 2000, a highly credible person stated “I am told one member of the Mega Society is a billionaire.” It only took $725 million to rank among the 400 richest Americans, but his person, if he exists, probably never appeared in Forbes, but just as different IQ tests don’t correlate perfectly, different ways of assessing and identifying wealth don’t correlate perfectly. But given how few people have qualified for the Mega society, it would be pretty fascinating if one of them was a USD billionaire by any credible measure.
Before I get to part 3 of my series high range tests, I wanted to quickly share my new theory on autism (or at least I think it’s new; maybe someone had similar ideas before).
When we are babies our brain begins to form a huge number of connections between neurons to help humans adapt to any environment we might be raised in. By our teens, it’s clear what type of environment we’re in and the brain begins to prune any connections that are not needed in that environment, to preserve resources for the needed ones.
However autistic kids show considerably less pruning than their peers.
Here’s where my theory comes in: autistics are thought to lack social and practical intelligence, or common sense. But common sense comes from experience and so we should expect autistics to be more impaired in acquiring experience based skills because they lack the very pruning process that diverts resources to the neural connections that experience tells them they need.
Does that make them less intelligent? I define intelligence as your ability to adapt, so on the one hand, not having the pruning process makes them much less adapted to their environment, because pruning evolved to maximize the functioning of specifically those connections needed in your environment.
On the other hand, if the environment rapidly changes, the autistic brain will have all these extra synapses ready to exploit it. This may explain why allegedly autistic types like Bill Gates were able to dominate when technology changed so rapidly, the pruning process could not keep up with it. It also explains why autistics tend to do better on so-called fluid tests of novel problem solving than on crystallized tests of acquired knowledge and why autistics seem to have a child-like personality (children like autists, have too many synapses).
So the autistic mind, like the mind of a child, is both adaptable and unadapted: unadapted to the environment they were raised in. Adaptable when the environment changes.
Now schizophrenia is sometimes said to be the opposite of autism, and indeed, schizophrenics show the opposite pattern: too much synaptic pruning. This would result in the social common sense part of the brain becoming too strong at the expense of the ability to learn new stuff. So common sense will tell them, there must be a conspiracy, but when no evidence of a conspiracy occurs, they can’t revise their theories. By contrast autistics may never grasp the common sense theory, but will constantly come up with new ones. This also explains why schizophrenics have higher rates of dementia: too much pruning means less cognitive reserve in old age when the brain naturally shrinks.
As we saw in in part 1, a reader had some questions about high range power tests.
The reader asks:
What, if any, capacities might tests of the high-range “power” format disclose that standard IQ tests cannot, or, at least, do not?
As I mentioned in part 1, they probably measure the personality trait TIE (Typical Intellectual Engagement) and perhaps some cognitive abilities that conventional IQ tests miss like executive function. And as discussed in the comment section, they’re probably less sensitive to test anxiety because you can take them in a relaxed non-threatening environment.
But there’s more.
Creativity
Chris Langan stated:
Certain high-ceiling intelligence tests, generically called “power tests”, are composed of extremely difficult items requiring higher levels of problem-solving ability than the items on ordinary IQ tests. Since these items usually have no known algorithms, their solutions cannot be looked up in a textbook, and where subjects do not know each other, one must rely on intrinsic problem solving ability.
From Discussions on Genius and Intelligence Mega Foundation Interview with Arthur Jensen pg 23
Arthur Jensen replied:
…Solving problems, or even thinking up problems, for which there are presently no algorithms, takes us into the realm of the nature of creativity. There are as yet no psychometric tests for creativity in a nontrivial sense. We can’t (yet) predict creativity or measure it as an individual trait, but can only examine its products after the fact.
From Discussions on Genius and Intelligence Mega Foundation Interview with Arthur Jensen pg 24
I find Jensen’s reply curious. He just admitted that the type of psychometric test Langan was describing involved creativity and then denied any tests measure it. Although I was extremely impressed by the questions Langan asked, he should have asked for clarification here.
The so-called distinction between creativity and intelligence is interesting. Intelligence is commonly defined as your ability to problem solve, but what is creativity if not original solutions to problems? So I guess people deny conventional IQ tests measure creativity because the solutions aren’t original enough. Why don’t conventional IQ tests require original solutions? Because such solutions would be so numerous that the test scoring manual could not include them all, or if there’s only one, in order for it to be original, too few people would discover it, making it useless for mass testing.
But because untimed power tests include many problems very few people can solve, by definition they measure original problem solving and thus creativity. One could claim that the problems must have social significance to be true measures of creativity, but what is significant is context dependent and creativity, like all traits, is relatively stable.
For example, before coronavirus became a global pandemic, creating a vaccine would have been unimportant but if someone had created one in 2018, they wouldn’t think “if I had only waited until 2020, I could have been creative, instead I’m merely extremely intelligent”. There could be a parallel universe where problems on the Mega Test have enormous real-world implications, while discovering relativity is merely a hard item on the Mega Test. Creative is something you either are or aren’t, it’s not something that changes with the social value a particular society puts on a given problem at a given time.
Autism?
Commenter “Mug of Pee” jealously goes ballistic when anyone values tests other than the ones he scored high on (SAT, GRE). In order to devalue such tests, he’s claimed, somewhat facetously, that the Mega Test measures autism. While it’s true that the Mega Test requires you to focus for very long periods of time (an autistic trait), it also requires you to be interested in a wide variety of subjects, as opposed to the narrow autistic focus. I suppose there could be some autistics whose area of obsession is just puzzle solving in general, but I know of no confirmed cases.
Without doing brain scans, autism is much harder to measure as objectively as IQ but if forced to do so, I would use one’s composite score on the following four variables:
income adjusted for IQ (the lower the more autistic)
occupational status adjusted for IQ (the lower the more autistic)
head size adjusted for IQ (the larger the more autistic)
theism adjusted for IQ (the higher, the less autistic)
In other words, autistics would tend to be those who are poorer, less respected, bigger brained and less theistic than their conventional IQs predict. Anecdotal evidence suggests people with high Mega Test scores would fit the first three criteria, but perhaps not the fourth. However I’m assuming a linear relationship between IQ and all these variables. If at the highest levels, IQ predicts “success” in a curvilinear way, we might find that the socio-economic underachievement of some Mega society members is not atypical of their IQs as measured by conventional tests (with high ceilings).
What are the real-world implications of an exceptional score (say, 170 – 190) on tests of the Mega Test type?
While conventional IQ tests are administered in a single sitting, and supervised so that you can’t rely on books or calculators to “cheat”, Mega Type tests include problems so complex, they’re solved over dozens of unsupervised hours in many sittings, with the use of reference books and other aids.
Chris Langan stated:
…by virtue of their difficulty, these problems take longer to solve… sometimes days or even weeks. Accordingly, power tests are untimed and unsupervised. This opens the door to factors like motivation and persistence, which are not among the factors primarily measured by standard IQ tests. On the other hand, virtually every significant intellectual achievement of mankind has involved these factors in great measure.
From Discussions on Genius and Intelligence Mega Foundation Interview with Arthur Jensen pg 23
Arthur Jensen stated:
…Such tests would have little practical use, although they could be of scientific interest in studying the nature of high-level problem solving. But people even capable of taking such tests could be identified with some conventional tests, such as a combination of the Advanced Raven Matrices and Terman’s Concept Mastery Test. People with high scores on such tests can demonstrate their problem solving ability in their careers. What is the need for prior selection? They can make it into college and graduate school if they’ve got high IQs, and it will be their virtually unique constellation of traits (g + special abilities + motivation + character, etc.) that will determine whether they will, first of all, identify important problems, and secondly, be able to solve them or at least materially contribute to their eventual solution.
From Discussions on Genius and Intelligence Mega Foundation Interview with Arthur Jensen pg 24
So Langan and Jensen seems to feel that Mega type tests measure a combination of traits useful for high level intellectual achievement. Cognitive traits (g, special abilities) + personality variables (persistence + motivation). Jensen seems to feel that the cognitive component can be sufficiently measured by combining conventional psychometric tests (eg RAPM + CMT), but what about the non-cognitive component?
Jensen writes about a trait called Typical Intellectual Engagement (or TIE) :
Most people perform at near their maximum level while taking a cognitive test. However, even among persons who show exactly the same level of g, there is great variation in TIE, which is assessed with a fifty-nine-item self-report questionnaire. The TIE inventory assesses the degree to which the individual typically engages in g-demanding activities, vocationally and especially avocationally, and has what would ordinarily be regarded as intellectual interests (reading, learning, thinking, a wide range of interests, particularly in literature, science, and mentally challenging activities, such as chess, being absorbed by the subjects of one’s interests and delving into them in depth).
The g Factor by Arthur Jensen pg 574
Jensen continues:
TIE is much more a personality factor than an ability factor. It does not correlate at all with G f (or a third-order factor g), but has significant but small correlations (r = + .3 to +.4) with verbal IQ and Gc. (Tests for IQ and Gc involve specific knowledge content and hence reflect intellectual achievement as well as the information-processing capacity that is measured more purely by Gf.) For a given level of g, a higher level of TIE in adulthood leads to somewhat higher levels of real-world intellectual achievement. But TIE itself clearly belongs in the personality domain, as shown by its correlations of about + .60 with two of the “ Big Five” personality factors (“ Openness” and “ Conscientiousness” ), as well as with another personality factor, “ self-directed activity,” which reflects energy level, absorption, and lack of distractibility.
The g Factor by Arthur Jensen pg 574
Given the 0.57 correlation between IQ and academic success, I’d expect someone with about + 5 sigma score on a conventional IQ test (top one in 3.48 million) to on average, have real world intellectual achievements of about +5(0.57) = + 2.85 (top one in 482). In other words, an adjunct professor who never gets tenure. That’s not to deny that some +5 sigmas achieve far more academically, perhaps even winning the Nobel Prize, but the overall average would be dragged down by all the +5 sigmas who don’t even graduate from college, either because they have more lucrative options, or because they don’t have the luxury.
I would expect, if we could get a representative (rather than self-selected) sample of people with +5 sigma Mega scores, they would be about as academically successful as their counterparts on conventional IQ tests. Because the Mega Test likely loads more on personality variables than conventional tests do (in particular TIE because you have to stay intellectually engaged with a problem for days or weeks), it likely has a lower g loading, but because TIE correlates with academic success independently of g, it’s correlation with real-world intellectual achievement is probably about the same.
I do wonder though whether personality traits like TIE are the only difference between people who perform better on Mega type tests than conventional tests. Could there be cognitive differences too? Conventional IQ tests break big problems down into digestible mini-problems but perhaps the ability to break big problems into small ones is an important cognitive ability in its own right. When faced with big problem in real life, even something as trivial as writing a blog post on a big topic, I’m sometimes overwhelmed with a sense of where do I even begin? Perhaps Mega type tests better assess this kind of big picture planning and hierarchical organizing. It would be interesting to see whether neuro-psychological tests of executive functioning predict Mega test performance independently of both conventional IQ and TIE.
This is the third and final article in a series ordered for $3.50 USD by 150 IQ Ganzir (the market has since driven the price up to $15). Here, I focus on a 1958 paper on 64 Yeshiva men (age 16 to 31; mean age 21.43). Author Boris M. Levinson states:
Our sample finally consisted of 64 subjects, classified as follows : (a) six senior Yeshiva High School students, (b) 31 Yeshiva College students, (c) 27 graduate students. Among them were four ordained rabbis. Every graduate student was an alumnus of Yeshiva College. The writer believes that the sample secured was fairly representative of the Yeshiva population…
Below are the scores of the sample on the original Wechsler Adult Intelligence Scale (WAIS). The 11 subtest scores are expressed on a scale where the average American scores 10 (SD = 3), and the verbal, performance, and full-scale IQs are expressed on a scale where the average American scores 100 (SD = 15). To convert scaled scores to IQ equivalents, just multiply by 5 and then add 50. Because the study, was submitted for publication in 1957 and the WAIS was standardized in 1953.5, there was a trivial Flynn effect I adjusted for.
Flynn effect adjustments were made by assuming the U.S. gains from the 1952.5 WAIS norming to the 1978 WAIS norming reported on page 240 of James Flynn’s Are We Getting Smarter? were linear over that period & thus calculating the expected gain in 1957 & subtracting it from the Yeshiva scores.
After correcting for slightly inflated norms, not a single mean score on the verbal subtests fell below the “Bright Normal range” and not a single performance subtest fell above the Average range. This complete lack of overlap shows the validity of the original verbal vs performance dichotomy that the Wechsler scales have since abandoned but which reflects the likely original reasons human intelligence evolved: To talk (verbal IQ) and to make and use tools (performance IQ).
Even Wechsler himself, who was also a Jewish New Yorker, likely showed this verbal > performance gap, as he became concerned that he could not solve his own Block Design items quickly.
The mean verbal > performance gap in this study is 21 points (22 after I adjust for norm inflation) but Levinson notes that for the graduate student sub-sample (n = 27), the mean gap was about 26 points and for rabbis (n = 4) it was about 34, which he interpreted as evidence that exposure to traditional Jewish culture was causing the gaps.
Levinson writes:
The students stated that the performance items appeared childish and unimportant. While they tried to achieve a good score whenever possible, the verbal items were more of a challenge...They said that an inferior score on a performance test was not as damaging to their self-esteem as a poor mark on the verbal tests.
I had the opposite attitude when I was tested on the Wechsler children’s scale at age 12. I thought the verbal items were mostly measuring what I learned in school and thus not knowing an answer didn’t damage my self-esteem. By contrast I saw the performance items as measuring my real intelligence.
Levinson writes:
In discussing the digit span test, quite a few of them indicated that they had visualized the digits. This is similar to the experience of some of these students who, in studying the Talmud, remember page location of a passage.
On the other hand, practice on one cognitive task seldom transfer to another unless they’re all but identical. In one study people who practiced their memory span for digits managed to increase it from seven to in some case over 100, but when faced with a memory span for letters task. They were right back down to seven. But perhaps that study was too brief to fully capture the effects of practice.
Levinson writes:
The subjects have been subjected, since their early school days, to a curriculum which greatly emphasized verbal knowledge, rote memory, verbal concept formation, abstract ideas, to the general neglect of performance arts. Examinations in the Talmud, for example, are oral and emphasize the detailed memorization and understanding of tracts. A differentiation of intellectual abilities has thus occurred. These cultural forces have also brought about different attitudes and self concepts regarding various intellectual tasks. Failure on a performance item does not carry the same ego deflating connotations as failures on a verbal task. It thus appears that the greater the premium placed on verbal ability in a subculture, the greater will be the disparity between verbal and performance WAIS scores.
There’s no doubt that cultural exposure affects verbal IQ score. Excellent research in the 1920s showed that canal boat children who lived a nomadic existence where they were virtually deprived of schooling, showed massive declines in IQ as they got older. Because IQ tests are normed for age, and because these kids were kept out of school they fell further and further behind their chronological age-mates on the type of knowledge that IQ tests measure. Young canal boat kids would have an IQ around 90, but older canal boat kids would have an IQ of 60. However when their performance IQs were tested, there was no such decline.
Perhaps the opposite is occurring with Yeshiva kids. Instead of their verbal IQs falling further and further behind as the effects of missed schooling accumulate, they grow further and further ahead, as the effects of enhanced schooling multiply.
On the other hand schooling likely has diminishing returns on verbal IQ once we get a basic amount. A study in Japan (where schooling is more intense) found older kids scored only slightly higher on verbal IQ.
