The Unz Review: An Alternative Media Selection
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 BlogviewJames Thompson Archive
Intelligence, Objectively
🔊 Listen RSS
Email This Page to Someone

 Remember My Information



=>

Bookmark Toggle AllToCAdd to LibraryRemove from Library • BShow CommentNext New CommentNext New ReplyRead More
ReplyAgree/Disagree/Etc. More... This Commenter This Thread Hide Thread Display All Comments
AgreeDisagreeLOLTroll
These buttons register your public Agreement, Disagreement, Troll, or LOL with the selected comment. They are ONLY available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also ONLY be used three times during any eight hour period.
Ignore Commenter Follow Commenter
Search Text Case Sensitive  Exact Words  Include Comments
List of Bookmarks

It takes a certain courage to title a paper: Genetic “General Intelligence,” Objectively Determined and Measured.

Javier de la Fuente, Gail Davies, Andrew D. Grotzinger, Elliot M. Tucker-Drob, Ian J. Deary

doi: https://doi.org/10.1101/766600

Objectively? Is such language permissible in contemporary science? Should we not instead be cautiously shuffling towards seven types of ambiguity, hedged in with eight layers of limitations? Who are these wild types willing to risk all in search of glory? In fact, a look at the names shows that this group have established an excellent track record, so they have in all probability chosen their words carefully, with the facts on their side.

First, a digression. Since 1969 there have been factions eager to denigrate intelligence, saying that its measures are based on arbitrary mental tasks, gathered together in the statistical artefact of a make-believe common factor, which is based on precisely nothing. How can one possible counter this all-encompassing dismissal of the psychometric project?

One approach would be to link a common factor to a genetic substrate, and anchor it in the genome.

It has been known for 115 years that, in humans, diverse cognitive traits are positively intercorrelated; this forms the basis for the general factor of intelligence (g). We directly test for a genetic basis for g using data from seven different cognitive tests (N = 11,263 to N = 331,679) and genome-wide autosomal single nucleotide polymorphisms. A genetic g factor accounts for 58.4% (SE = 4.8%) of the genetic variance in the cognitive traits, with trait-specific genetic factors accounting for the remaining 41.6%. We distill genetic loci broadly relevant for many cognitive traits (g) from loci associated with only individual cognitive traits. These results elucidate the etiological basis for a long-known yet poorly-understood phenomenon, revealing a fundamental dimension of genetic sharing across diverse cognitive traits.

The authors go on to explain that tests vary in the amount of general or specific factors required for their successful completion. Some variance in each test is shared with all other tests (g) and some is specific to each test (s). Hundreds of studies show that the g factor replicates, and accounts for 40% of test variance. Twin studies show that general intelligence is strongly heritable, suggesting an overlapping genetic architecture. However, the GWAS approach does not distinguish between “g” and “s”. The authors try to search for “g” directly, using a multivariate molecular genetics approach to the hierarchy of intelligence, g at the top, cognitive domains second, and individual tests at the bottom.

They used UK Biobank, blessed be its name, and seven cognitive tests:

Reaction Time (n = 330,024; perceptual motor speed), Matrix Pattern Recognition (n = 11,356; nonverbal reasoning), Verbal Numerical Reasoning (VNR; n = 171,304; verbal and numeric problem solving; the test is called ‘Fluid intelligence’ in UK Biobank), Symbol Digit Substitution (n = 87,741; information processing speed), Pairs Matching Test (n = 331,679; episodic memory), Tower Rearranging (n = 11,263; executive functioning), and Trail Making Test – B (Trails-B; n = 78,547; executive functioning). A positive manifold of phenotypic correlations was observed across the seven cognitive traits.

The authors then investigate the genetic contribution of g to variation in each of the cognitive tests. Genetic correlation is simply the correlation between the genetic contributors to each of the measured abilities. It is correlation at the level of genes, not test scores. If the brain is made up of modules, then one would expect such genetic correlations to be low. On the other hand, a brain largely based on general ability would have strong correlations. In fact, the genetic correlations range from .14 to .87, with a mean of .53 and the first principal component accounted for a total of 62.17% of the genetic variance. The genetics of intelligence is largely g based, it would seem.

Further work identifies the tests that are most g loaded:

Trails-B (95.30% genetic g; 4.70% genetic s), Tower (72.80% genetic g; 27.20% genetic s), Symbol Digit (69.10% genetic g; 30.90% genetic s), and Matrices (68.20% genetic g; 31.80% genetic s). Verbal Numerical Reasoning (51.40% genetic g; 48.60% genetic s) and Memory (42.40% genetic g; 57.60% genetic s) are more evenly split. Reaction Time has the majority of its genetic influence from a genetic s (9.50% genetic g; 90.50% genetic s). We emphasize one important implication of these results, i.e. that genetic analyses of some of these individual traits will largely reveal results relevant to g rather than to the specific abilities thought to be required to perform the test.

Reaction time is somewhat of an outlier from the genetic point of view, as might be expected by the very simple, knee-jerk nature of the task.

Anyway, which locations on the genome are contributors to g? Getting an answer is important, since a GWAS hit could be generalizable to a broad universe of cognitive traits, or specific to a particular task, and knowing which makes a difference. In the explanation below, Q is a measure of heterogeneity, opposite to g.

Miami plot of unique, independent hits for genetic g (top) and Q (bottom). Q is a heterogeneity statistic that indexes whether a SNP evinces patterns of associations with the cognitive traits that departs from the pattern that would be expected if it were to act on the traits via genetic g. The solid grey horizontal lines are the genome-wide significance threshold (p < 5×10−8) and the dotted grey horizontal lines are the suggestive threshold (p < 1 ×10−5). The following genome wide significant loci are highlighted: Red triangles : g loci unique of univariate loci. Blue triangles : g loci in common with univariate loci. Green circles : univariate loci not in common with g loci. Yellow triangles : g loci in common with Q loci. Yellow diamonds : Q loci unique of g loci.

Overall, we identified 30 genome-wide significant (p < 5×10−8) loci for genetic g, 23 of which were common with the univariate GWAS of the individual cognitive traits that served as the basis for our multivariate analysis. We identified, in total, 24 genome-wide significant loci for Q, 3 of which were significantly associated with genetic g (and therefore likely to be relevant to more specific cognitive traits, and false discoveries on g) and 15 of which were significantly associated with at least one individual cognitive trait in the test-specific GWASs.

Although it was not intended to be part of the study, seven new locations for memory were found, and some of those locations have been associated with Schizophrenia, anti-saccade response, linguistic errors, hand grip strength and bone mineral density.

So, have the authors “got away with” their combative title? The best way to answer would be to set the question “What else do you want?” The claim is that intelligence is real, and is a real aspect of the brain. To show that that is the case you can show a link between intelligence test scores and real life (this has been done many, many times, and some examples are shown below) and a link between intelligence test scores and implied measures of genetic heritability via twin and family studies (also done many times), and now finally a link between intelligence test scores broken up into general and specific factors and measures of heritability via actual genomic studies identifying locations for general and specific factors of intelligence.

Here are some correlations between intelligence and real life measures
https://archive.ph/PCvgk

In my view this is a very important advance. It shows an underlying reality, at a genetic level, between general and specific aspects of cognitive ability. It allows investigations to proceed at two levels: the test score level and the genomic code level. Further studies will drill down into yet more detail.

It is fair to say that this is an objective approach, and ought to answer any reasonable critic of the reality of cognitive ability being based on brains which are under substantial genetic control.

 
• Category: Science • Tags: Behavior Genetics, General Intelligence, IQ 
Hide 239 CommentsLeave a Comment
Commenters to Ignore...to FollowEndorsed Only
Trim Comments?
    []
  1. Spearman would be so proud! His work from 115 & a half years ago has aged extremely well:) & now with bonus genetic info! but how many more years will it take the public to understand?
    “General Intelligence,” Objectively Determined and Measured (C. Spearman, 1904)
    https://www.jstor.org/stable/1412107?seq=1#metadata_info_tab_contents

    • Replies: @James Thompson
  2. res says:

    Interesting paper and nice summary. Thanks!

    It is too bad they did not separate reaction time into response and movement time as Rushton and Jensen (2005) discusses: https://www1.udel.edu/educ/gottfredson/30years/Rushton-Jensen30years.pdf

    The response times significantly correlated (negatively) with Raven Matrices scores, whereas movement times have a near-zero correlation. The average reaction times for the three racial groups differ significantly (see Figure 2). They cannot be explained by the groups’ differences in motivation because the East Asian children averaged a shorter response time but a longer movement time than did the Black children.

    • Replies: @James Thompson
    , @Aft
  3. Inclusion or exclusion of items from tests is subjective, not objective. Items on IQ tests, to quote Jensen (1980: 148) “emerge arbitrarily from the heads of test constructors.” Therefore ,”intelligence” (IQ test scores) are subjective—not objective—measures. James Thompson’s title “Intelligence, objectively” is false on the basis that there is no underlying theory of item inclusion and exclusion. Items on the Raven, for example, are chosen on the basis of intuition, not on the basis of an underlying processing theory (see Carpenter, Just and Shell, 1990).

    Elaine Castles (2013: 29) in her book Inventing Intelligence notes that “Intelligence is a cultural construct, specific to a certain time and place.” Castles (2013: 152) further states that “to assert that even the most carefully constructed [IQ] test is somehow a pure measure of hypothetical intelligence flies in the face of both scientific evidence and common sense.” This is because “g” itself isn’t a real physiological entity, it’s built into the test through test construction (Richardson, 2002).

    This, of course, then gels with Bronfenbrenner’s and Ceci’s bio-ecological model of human development (and IQ). Large-scale cultural differences, therefore, contribute to psychological differences between groups. The way one thinks about something is influenced by their culture and their interactions with other individuals in that culture. Meaning, in a cultural context, is always between individuals within that culture which will therefore shape culture and therefore thought. This, though, is a foregone conclusion, as Prinz (2014: 67) notes in his book Beyond Human Nature: How Culture and Experience Shape the Human Mind that “cultural differences can have an impact on psychological traits.

    Nevermind the fact that no IQ test is construct valid – that is, there is no theory of individual intelligence differences so we can’t logically state that what IQ tests are measuring is truly “intelligence.” A logical analysis of the items on the test should lead one to believe that IQ tests tests specific cultural—class—knowledge and skills. For instance, “IQ” stands for intelligence quotient; intelligence is cognition; the score one gets on an IQ test is due to thinking; analyzing the items and structure of the test leads one to the conclusion that the items are class-specific (culture-specific); therefore, “intelligence quotient” (IQ) test scores are, in effect, “thinking scores”—one’s thinking score in a specific, in this case middle-class, cultural context.

    The “common factor” that emerges is due to sociocognitive-affective measures which influences one’s performance on the test, such as differing knowledge gained due to one’s membership in a certain social class or culture; variation in self-efficacy beliefs; and variation in test anxiety, self-confidence, and motivation etc affect one’s performance (Richardson, 2002: 288).

    In all, the IQ-ist racket is exposed and has been exposed for decades—items on IQ tests are biased, culturally and class-wise; there is no construct validity for any IQ test so psychometricians have no idea what they’re measuring; and there is no theory of individual intelligence differences nor human intelligence. These problems and more point to the conclusion that IQ tests are no testing “intelligence”, but class-specific knowledge and skills, not “general intelligence”.

  4. @res

    Agree that a breakdown of movement and thinking time would be good, but I don’t think it was collected. Interestingly, in their work on ageing they found that simple reaction time was more predictive of ageing than was choice reaction time, which is somewhat against prediction. Jensen and Rushton’s analysis is very elegant, but I don’t know if anyone has tried to replicate it.

    • Replies: @Aft
  5. @egregious philbin

    Spearman vindicated. As Ian Deary says, “All psychometry is a footnote to Spearman”.

  6. Realist says:
    @RaceRealist88

    Spin it the way you want but there are intelligence differences between races.

    • Replies: @Anonymous
  7. @RaceRealist88

    Inclusion or exclusion of items from tests is subjective, not objective.

    Ugh. That’s like saying that using a 40-yard sprint time to test running speed is invalid because you could have “subjectively” picked 100 yards or 25 yards or 26 miles as the measuring distance.

    It doesn’t matter how you pick your test questions. The point is that the the test result will be valid if it: (a) is internally valid (i.e. it tends to return the same value for the same individual over multiple testings); and (b) is externally valid (i.e., the results of the people who take the test correlate well with the objectively measurable external reality you are interested in).

    In the case of IQ the external reality is generalized problem-solving ability, academic success, etc. If you want to propose a different set of questions or a different external measure of intelligence to correlate your test against, knock yourself out. If your test results are internally and externally valid, then congratulation! your test “works.”

    It sounds plausible to suggest that IQ results are dependent on “cultural.” But that is easily tested. If “culture-specific” tests produced results that better satisfied the internal and external validity criteria you’d have point. But this doesn’t work. (Except for obvious things like not giving a test in a language the subject doesn’t understand).

    The other problem with “the IQ as culture” nonsense is that “it proves too much.” IQ tests correlate far better to measurable reality than any other form of psychological testing. So if you refuse to accept the validity of IQ tests, you would also be compelled to throw out the entire discipline of psychology as pseudo-science as well.

  8. @Hypnotoad666

    “That’s like saying that using a 40-yard sprint time to test running speed is invalid because you could have “subjectively” picked 100 yards or 25 yards or 26 miles as the measuring distance.”

    False. Athletic ability and IQ are completely different.

    https://notpoliticallycorrect.me/2017/12/30/athletic-ability-and-iq/

    “The point is that the the test result will be valid if it”

    A test is valid if and only if there is a theory for what is being measured.
    There is no theory of what is being measured on IQ tests.
    Therefore, IQ tests aren’t valid.

    “It sounds plausible to suggest that IQ results are dependent on “cultural.” But that is easily tested. If “culture-specific” tests produced results that better satisfied the internal and external validity criteria you’d have point. But this doesn’t work. (Except for obvious things like not giving a test in a language the subject doesn’t understand).

    What kinds of items are on IQ tests? I have quite a few examples—do you have any to back your assertions?

    “The other problem with “the IQ as culture” nonsense is that “it proves too much.” IQ tests correlate far better to measurable reality than any other form of psychological testing. So if you refuse to accept the validity of IQ tests, you would also be compelled to throw out the entire discipline of psychology as pseudo-science as well.”

    No problem throwing out all of psychology as pseudoscience—you’re appealing to consequences, in any case. In virtue of what is the claim that “IQ testing” (and all tests of ability) are culture-bound? If a West African Binet constructed a so-called culture-fair test, would it be fair? Why or why not?

  9. @RaceRealist88

    IQ tests are no[t] testing “intelligence”, but class-specific knowledge and skills, not “general intelligence”.

    If a West African Binet constructed

    What kinds of items are on IQ tests? I have quite a few examples

    Would “regatta” be one of those items?

    you’re appealing to consequences

    >implies that “appealing to consequences” is a cardinal epistemological error

    >in a discussion about test construction

    I think we’ve found a West African genius! Or a (((West African))) one…

    • Replies: @Hippopotamusdrome
  10. utu says:

    As often in papers like this one what is important is obfuscated. So, what is important? How well PGS (polygenic score) can predict g or their “genetic g”? Look for the paragraph that begins with:

    We next created polygenic scores (PGSs) for genetic g and the individual UK Biobank cognitive traits and used them to individually and simultaneously predict variance in cognitive performance and educational attainment in an independent study…

    Can you find variance predicted/explained? No, it is nowhere to be found, including the supplementary material. But they proudly talk about improvements:

    We found that the proportional increase in R2 in Digit Symbol by the Symbol Digit PGS beyond the genetic g PGS was <1%, whereas the genetic g PGS improved polygenic prediction beyond the Symbol Digit PGS by over 100%…

    or

    g PGS alone for those Generation Scotland phenotypes most directly related to verbal knowledge: Mill Hill Vocabulary (62.45% increase) and Educational Attainment (72.59%).

    It seems that they are very careful not to reveal what the actual value of variance explained was. That it increased 62,45% does not tell us much if it was still below 9-11%.

    Basically they are dishonest fuckers who are guilty of obfuscation at least.

    • Replies: @res
  11. res says:
    @utu

    Can you find variance predicted/explained? No, it is nowhere to be found, including the supplementary material.

    It seems that they are very careful not to reveal what the actual value of variance explained was.

    I think you failed to scroll through all of the sheets in the Supplemental Material (my copy of the spreadsheet opened on the last sheet). From the paper:

    We next created polygenic scores (PGSs) for genetic g and the individual UK Biobank cognitive traits and used them to individually and simultaneously predict variance in cognitive performance and educational attainment in an independent study, Generation Scotland; N=6,950 unrelated individuals (Table S7).

    Looking at Table S7 (the first sheet), the numbers you want are there. Now feel free to complain about how small the various R^2 are. Which would at least be a reasonable complaint.

    P.S. If you are going to use language like in that first quote (and especially in your final sentence which I did not quote), it is wise to double check your correctness before commenting.

    • Replies: @utu
  12. szopen says:
    @RaceRealist88

    So you are basically arguing that tests are not for intelligence, but only for social class skills; and it’s not intelligence which is genetically based, but traits which help you in modern world to advance to middle class (or to fall below; because IQ tests are different within families). In other words, you don’t believe intelligence is heritable. Instead you surely must believe that class status is heritable (i.e. it’s not that you are born intelligent; it’s just you are born with set of traits which will doom you to fall to lower class, or to raise to high class).

    Not sure why the distinction matters (I’d say traits required one to belong to middle class are rough proxy of intelligence).

  13. szopen says:
    @RaceRealist88

    I say you are using overly strict criteria.

    For example, you are saying “A test is valid if and only if there is a theory for what is being measured.” (i.e. “a test is valid only if it satisfies my criterion, and not when it satisfies your criterion” – that means that your use of “test is valid” is different from someone else’s use of “test is valid”). But what it means “a a theory of what is measured”? If some barbarians measure weight by using primitive instruments and literal stones, but have no idea why some items are heavier than other and can’t say what is a weight exactly, does that mean their tests of weight do not reflect reality and are not at least some proxy of weight?

    It’s like trying to create tests for health. We all have some vague, imprecise sense of which people are “healthy” and who are not (just as our sense of what exactly is intelligence is rather imprecise). Before modern times, we could try to create some rough tests for who is healthy and who is not. They would not be “valid” in your picked sense, by your picked criteria; but if independent doctors, applying the same test, would get at least roughly similar tests and those tests would be useful for determining if a person needs medical attention, they would be useful; and while it would be OK to try to introduce some precise, scientific terms for things which contribute to common-folk “health” concept (to make things more clear and precise), when instead someone would just say “health” does not exist and dismissing everything with “health test are not valid” does not seem to me a well-thought position (i.e. when instead of making things better, someone just dismisses the whole concept).

    If a West African Binet constructed a so-called culture-fair test, would it be fair?

    What sense of fair? If it would predict, on average, equally good the outcomes in West-African society for whites, blacks etc (i.e. the result “100” would predict same outcomes for whites, blacks…) then IMO it would be fair.

    • Replies: @RaceRealist88
  14. @RaceRealist88

    Why aren’t Ravens Matrices culture fair as long as those takng them have had plenty of practice?

  15. mikemikev says:
    @RaceRealist88

    Why do people like you call yourselves the opposite of what you are? Is it a genetic tendency?

    • Agree: Gordo
  16. @RaceRealist88

    Agreed, but without any measurement, one sees that Africans, as a group, don’t care for any culture & learning, while east Asians are great zombie learners with rare flashes of originality. North Europeans are better in innovation than south Europeans, but frequently lack elementary survival instinct & are absurdly naive.

    Psychometry is great for measuring one’s abilities in various areas (numbers, space orientation, technical skills, words & vocabulary,..), but it cannot measure strong creativity, focus, will, work ethic, mental illnesses that hamper one’s life etc.

    • Replies: @Aft
  17. Does the above show that memory is less genetically influenced than most the other stuff?

  18. eugyppius says:
    @RaceRealist88

    Some of your references are actually near at hand to me right now, so a few points:

    You misrepresent Jensen. He is talking specifically about items in “tests of general information,” not IQ tests broadly. Jensen’s conclusion is absolutely not that “IQ test scores” are “subjective”, and not even that information (sub)tests are subjective.

    Carpenter, Just and Shell (1990) say the following about intuition and the Raven:

    [John Raven] wanted to develop a series of overlapping, homogeneous problems whose solutions required different abilities. However, the descriptions of the abilities that Raven intended to measure are primarily characteristics of the problems, not specifications of the requisite cognitive processes. John Raven constructed problems that focused on each of six different problem characteristics, which approximately correspond to the different types of rules that we describe later. He used his intuition and clinical experience to rank order the difficulty of the six problem types. Many years later, normative data […] became the basis for selecting problems for retention in newer versions of the test and for arranging the problems in order of increasing difficulty, without regard to any underlying processing theory. Thus, the version of the test that is examined in this research is an amalgam of Raven’s implicit theory of the components of reasoning ability and subsequent item selection and ordering done on an actuarial basis.

    From this we learn three things, all of which you have obscured or misrepresented in a very short space: 1) Raven’s original items corresponded directly to the mental abilities he wanted to measure (i.e., quantitative pairwise progression, figure addition or subtraction, distribution of 2-3 values, etc.). 2) Raven used his “intuition” to rank the difficulty of the items. 3) The relative difficulty of items on modern Ravens matrices tests is established according to “normative data.” One struggles to see what room is left for intuition here. Also you could perhaps read to the end of this article (it is not very long) to encounter some of the theories of intelligence and abstract reasoning that you claim over and over do not exist.

    You do seem to have digested the thoughts of silly people and to represent them correctly, however, which is something. People like Castles, who inter alia uses Gould’s assault on g to (try to) take down Spearman; and people like Richardson, who is so afraid of IQ testing that he wants to ban it, but who also believes IQ tests measure nothing clear and predict mostly results on similar tests. God isn’t that Galileo guy an idiot? Also please keep him away from all optical equipment for the love of all that is holy.

    Large-scale cultural differences, therefore, contribute to psychological differences between groups.

    Where do large-scale cultural differences come from?

    The “common factor” that emerges is due to sociocognitive-affective measures which influences one’s performance on the test, such as differing knowledge gained due to one’s membership in a certain social class or culture;

    Then why are precisely culturally specific (or culturally more specific) items less “g-loaded” (i.e., less subject to correlation) than items that are not?

    there is no theory of individual intelligence differences

    Uh, this is not true?

    • Agree: Aft
    • Replies: @RaceRealist88
  19. gotmituns says:

    Intelligence, Objectively
    ———————————-
    Just another dumb article about intellect. The only intelligence that matters anymore is the ability to shoot straight and hit your target. Everything else is secondary to that.

  20. Aft says:
    @James Thompson

    First, very cool paper and great summary.

    Re: simple reaction time, very curious that it’s so phenotypically correlated with g but not as genetically correlated, which would both point to it being more driven by the (unshared within developed countries) environment influences on intelligence: random prenatal events, lifetime immune load and toxin burden etc.

    The fact that it’s also more correlated with aging than more g-loaded complex reaction times would seem to further strength that theory: reaction time may be far more sensitive to those accumulated stressors that also to a lesser degree impact g.

  21. Anonymous[940] • Disclaimer says:
    @Realist

    His is the default attitude for high-IQ people of low-IQ races (in his case, 90-IQ Southern Italy).
    Still better than high-IQ narcissists of high-IQ races who are denialists.

  22. Aft says:
    @RaceRealist88

    Peripherally related to some of the less intelligent comments (re: variance explained):

    One of the more interesting aspects of the recent large GWAS for intelligence is they’re a lot closer to accurately measuring additive genetic intelligence–the only portion you pass on to your offspring–than people think.

    12% of variance explained in the best study is a correlation of ~0.35 with phenotypic intelligence. That sounds small until you realize that phenotypic intelligence is itself only ~0.7 correlated with additive genetic intelligence (50% to 55% of variance explained is the consensus range for additive effects, with around 25% from dominance and 20% from unshared environment).

    Thus the actual correlation between currently available polygenic scores and one’s additive genetic intelligence is closer to 0.35/0.7 or 0.5. And that’s before taking into account measurement error, after which it’s above 0.5. Not as high as the 0.7 correlation between tested IQ and your additive genetics, but getting close.

    • Replies: @res
  23. Aft says:
    @RaceRealist88

    1. You are not a race realist. You’re clearly one of those verbally pseudo-skilled bullshit spewers saying things that look smart but are dead wrong and amply disproven.

    2. “If a West African Binet constructed a so-called culture-fair test, would it be fair? Why or why not?” As long as it involves a written non-physical test (no twerking sub-batteries) and shapes and colors like the Raven’s or block design, bring it on. We have a pretty good idea what the results will look like though.

  24. Realist says:

    Still better than high-IQ narcissists of high-IQ races who are denialists.

    Explain, please.

  25. Realist says:
    @RaceRealist88

    Here is an intelligence test that is accurate and simple: Make a list of the scientific advancements of the different races…the one with the most wins.

  26. eugyppius says:
    @RaceRealist88

    In for a penny, in for a pound: On the one hand I regret my earlier response to you after investigating your website and comment history. I see a few possibilities, among them that you’re running some kind of pro-IQ testing and/or pro-hereditarian (it is not always easy to tell) psyop; that you’re an internet troll whose race realism ceased to be an irritant after most of the internet cottoned on to the nature of the evidence; a very dimwitted person; or a mixture of all three.

    However that may be, I have followed your use of the “construct validity” mantra back for a full year now, and find it quite amusing.

    Nevermind the fact that no IQ test is construct valid – that is, there is no theory of individual intelligence differences so we can’t logically state that what IQ tests are measuring is truly “intelligence.”

    In past comments you write even more directly of the need for a mechanistic understanding of the relationship between tested IQ and intelligence for construct validation to be achieved. You write in another comment that the demanded theory has to be “well-accepted”, which is especially peculiar. Do you think that many measurements and observations about plants in botany from a hereditarian perspective would have been prima facie impossible in Stalinist Russia, when Lysenkoism prevailed? For that matter how can anybody measure or discuss a newly discovered phenomenon, when the theory underlying these measurements has not yet been published, let alone “accepted”?

    Anyway those are side matters. All of the research since Spearman surrounding the general factor of intelligence is construct validation. Via factor analysis it became possible to determine the degree to which this or that test can be said to measure g (even if you only assume the reality of general intelligence for the sake of argument).

    In yet another of your past comments, you write of the thermometer as being “validated,” by which you appear to mean that its measurements can be determined as accurate or not independent of its reading; or perhaps (being a little more generous to you) that there is a “well-accepted” theory explaining the accuracy and operation of thermometers.

    Here, however, you stumble very seriously for someone who has typed the words “construct validity” as many times as you have. For construct validation has nothing to do with thermometers. Do not take it from me, but from the people who coined the term in 1955:

    Construct validation is involved whenever a test is to be interpreted as a measure of some attribute or quality which is not “operationally defined.” The problem faced by the investigator is, “What constructs account for variance in test performance?” Construct validity calls for no new scientific approach. Much current research on tests of personality (9) is construct validation, usually without the benefit of a clear formulation of this process. Construct validity is not to be identified solely by particular investigative procedures, but by the orientation of the investigator. Criterion-oriented validity, as Bechtoldt emphasizes (3, p. 1245), “involves the acceptance of a set of operations as an adequate definition of whatever is to be measured.” When an investigator believes that no criterion available to him is fully valid, he perforce becomes interested in construct validity because this is the only way to avoid the “infinite frustration” of relating every criterion to some more ultimate standard (21). In content validation, acceptance of the universe of content as defining the variable to be measured is essential. Construct validity must be investigated whenever no criterion or universe of content is accepted as entirely adequate to define the quality to be measured.

    You can read the whole thing yourself and as a devotee of valid construction you probably should:

    http://psychclassics.yorku.ca/Cronbach/construct.htm#f1

    The thermometer as an index for temperature has what Cronbach/Meehl would call concurrent (not construct) validity. Intellectual acumen is a wholly different phenomenon in that no single measure or test can establish it. Rather, some tests appear to come closer (in that they are more predictive of life outcomes and also more tightly correlated with one another) than others. Via factor analysis clever people can nevertheless show what portion of any given test/measurement can be ascribed to the suspected phenomenon under study. Mechanistic theories of how intelligence functions are not necessary to this operation at all.

    • Replies: @eugyppius
  27. Anonymous[137] • Disclaimer says:

    The claim is that intelligence is real, and is a real aspect of the brain.

    No shit. It’s kind of embarrassing that this statement is “controversial” in certain “academic” circles. Same goes for genetic and heritability influence on the IQ. In our Western Clown World, every organ in an organism can be shaped by genes – except a human brain – which is somehow conjured out of the void, by a Social Justice fairy, to be exactly the same across the board.

    So, we’re still allowed to notice that individual dogs – or even dog species/breeds – are smart(er), but that’s unacceptable sacrilege when describing humans. Laughable.

    • Agree: SafeNow, Godfree Roberts
  28. eugyppius says:
    @eugyppius

    Just a follow-up as I procrastinate at the end of the day to illustrate how far RaceRealist88 has taken his terminological bone from whatever field he found it in.

    Cronbach/Meehl, as cited above; their humor is ponderous:

    Example of construct validation procedure. Suppose measure X correlates .50 with Y, the amount of palmar sweating induced when we tell a student that he has failed a Psychology I exam. Predictive validity of X for Y is adequately described by the coefficient, and a statement of the experimental and sampling conditions. If someone were to ask, “Isn’t there perhaps another way to interpret this correlation?” or “What other kinds of evidence can you bring to support your interpretation?”, we would hardly understand what he was asking because no interpretation has been made. These questions become relevant when the correlation is advanced as evidence that “test X measures anxiety proneness.” Alternative interpretations are possible; e.g., perhaps the test measures “academic aspiration,” in which case we will expect different results if we induce palmar sweating by economic threat. It is then reasonable to inquire about other kinds of evidence.

    The very scholars who developed the term “construct validation” are mysteriously unbothered by mechanistic generally accepted theories tying palm sweating to anxiety proneness.

    Add these facts from further studies: Test X correlates .45 with fraternity brothers’ ratings on “tenseness.” Test X correlates .55 with amount of intellectual inefficiency induced by painful electric shock, and .68 with the Taylor Anxiety scale. Mean X score decreases among four diagnosed groups in this order: anxiety state, reactive depression, “normal,” and psychopathic personality. And palmar sweat under threat of failure in Psychology I correlates .60 with threat of failure in mathematics. Negative results eliminate competing explanations of the X score; thus, findings of negligible correlations between X and social class, vocational aim, and value-orientation make it fairly safe to reject the suggestion that X measures “academic aspiration.” We can have substantial confidence that X does measure anxiety proneness if the current theory of anxiety can embrace the variates which yield positive correlations, and does not predict correlations where we found none.

    Here RaceRealist88 will probably think he has something to support his position at the point of reading about “the current theory of anxiety,” but it does not matter at all how anxiety is theorized or defined, only how the relationship as formulated is supported (or not) by tests that plumb alternate hypotheses.

    These final statements from just a little further on would, in a just world, lead RaceRealist88 to embrace a new mantra:

    At this point we should indicate summarily what we mean by a construct […] A construct is some postulated attribute of people, assumed to be reflected in test performance.

    What? It is not a mechanistically defined generally accepted theory?

    In test validation the attribute about which we make statements in interpreting a test is a construct. We expect a person at any time to possess or not possess a qualitative attribute (amnesia) or structure, or to possess some degree of a quantitative attribute (cheerfulness).

    All the time, one fears, the testers might lack precise mechanistic well-accepted theories of e.g. amnesia or cheerfulness. Strangely the developers of this cornerstone concept, so fatal to psychometric research, seem untroubled by this obstacle.

    A construct has certain associated meanings carried in statements of this general character: Persons who possess this attribute will, in situation X, act in manner Y (with a stated probability). The logic of construct validation is invoked whether the construct is highly systematized or loose, used in ramified theory or a few simple propositions, used in absolute prepositions or probability statements.

    It’s almost like a construct is a hypothesis explaining test results via some aspect of the test taker, and the validity of this construct can then be tested by further tests of alternate explanations. It’s almost like psychometric researchers have been doing that for like a century in their work on the general factor of intelligence.

    • Replies: @res
  29. Bill says:
    @RaceRealist88

    False. Athletic ability and IQ are completely different.

    As is typical with this guy, he provides no argument or evidence. If you follow his link, you will find no argument or evidence, just more bald conclusory assertions:

    Tests of athletic ability do not have any arbitrary judgments as IQ tests do in their construction and analysis of the items to be put on the test. It’s a simple, cut-and-dry explanation: on this instance in this test, runner X was better than runner Y.

    He simply 1) denies that arbitrary choices in measuring athletic performance are arbitrary 2) affirms that arbitrary choices in measuring intellectual performance are arbitrary and 3) affirms that this arbitrariness somehow invalidates measurement. He does this at some, tedious length, but that’s all there is.

    The interesting question is is RaceRealist88 really dumb enough to believe his own bullshit?

    • Agree: Aft
  30. @mikemikev

    It’s called crypsis, a form of stealth parasitism by deception and impersonation. (((Certain people))) are virtuosi at it.

  31. CanSpeccy says: • Website

    Since 1969 there have been factions eager to denigrate intelligence, saying that its measures are based on arbitrary mental tasks, gathered together in the statistical artefact of a make-believe common factor, which is based on precisely nothing. How can one possible counter this all-encompassing dismissal of the psychometric project?

    I don’t know if you are deliberately engaging in a logical sleight of hand, but surely no faction, let alone a number of them, is or ever has been “eager to denigrate intelligence”.

    From what you say following that remarkable initial statement seems to indicate that the eager factions of which you speak denigrate not intelligence, but the claims of those, such as yourself, who boast of an ability to measure intelligence on a single linear scale, a claim incidentally that to many will seem like a real denigration of intelligence.

  32. res says:
    @Aft

    Peripherally related to some of the less intelligent comments (re: variance explained):

    Not sure if you are including my comment 11 in that. I was amazed just how low the R^2 in Table S7 were (0-5%). This was out of sample prediction so that likely reduces them relative to most quoted figures, but still. Does anyone have a good explanation for why this genetic g only explains 4% of the variance for g in this study? This seems low to me for a current study based on the UKBB.

    And speaking of variance explained. Some might remember that one of my hobby horses is the possibility that naive correction for Principal Components (in an attempt to correct for population stratification) might correct away true genetic influence on a trait. In particular, if there are systematic causal population genetic differences for a phenotype then those will be corrected away (at minimum the coefficients will be reduced).

    This concern causes me to wonder why GWAS never (feel free to provide counterexamples) look at how much variance is explained by the PC (nice double entendre ; ) corrections.

    I finally ran across a GWAS which at least talks about this (emphasis mine).
    https://doi.org/10.1017/thg.2016.85

    Abstract
    We identified the genetic variants for eye color by Genome-Wide Association Study (GWAS) in a Dutch Caucasian family-based population sample and examined the genetic correlation between hair and eye color using data from unrelated participants from the Netherlands Twin Register. With the Genome-wide Complex Trait Analysis software package, we found strong genetic correlations between various combinations of hair and eye colors. The strongest positive correlations were found for blue eyes with blond hair (0.87) and brown eyes with dark hair (0.71), whereas blue eyes with dark hair and brown eyes with blond hair showed the strongest negative correlations (-0.64 and -0.94, respectively). Red hair with green/hazel eyes showed the weakest correlation (-0.14). All analyses were corrected for age and sex, and we explored the effects of correcting for principal components (PCs) that represent ancestry and describe the genetic stratification of the Netherlands. When including the first three PCs as covariates, the genetic correlations between the phenotypes disappeared. This is not unexpected since hair and eye colors strongly indicate the ancestry of an individual. This makes it difficult to separate the effects of population stratification and the true genetic effects of variants on these particular phenotypes.

    This was for only three PCs of a fairly genetically homogenous population. I would expect this problem to be even worse using more PCs for a more heterogenous population.

    Any thoughts?

    • Replies: @Aft
  33. res says:
    @eugyppius

    Thanks for your thorough exposition on this!

    • Replies: @eugyppius
  34. JoannF says:

    A fun article with a few hilarious reactions.
    Let me just sum up all that the egalitarian sectarians among us find unholy about testing “intelligence,” with the evil intent of assigning different levels of such to different homo subspecies which we deem not to exist.

    Obviously, intelligence is a non-genetic social construct – primatologists have come to the same conclusions.
    Bonobos would already have taken their place in “human” society if it wasn’t for racial and social prejudice, and if it wasn’t for the nasty fact that there seem to be genetic differences in grammar – grammar not being “intelligence” (a social construct) mind you, but rather a hardwired neurological phenomenon.
    Sort of athletic (I hope you read all the comments thus far).
    Seems that grammar still sets us apart, but nothing else does apparently, expect social oppression and athletic abilities, which are – as opposed to “intelligence” – as real as grammar.

    That Bonobos, natives of Africa, never invented written language (don’t try – they do have language, just, due to their more primitive grammar and larynx, not quite as complex ones as humans, which again are primitive enough as is) is however nothing to do with any lesser genetic ability, but rather with their primitive society, which is directly to blame on colonialism/capitalism and indirectly that higher evil, White Supremacy.
    The success of White Supremacy did not, (could not logically have) come to be by greater inventiveness through genetic higher genetic abilities but rather through the brutal distillation of Greater Greed which is somehow inherent to the The White Race, deeply socially entrenched by The Patriarchate, a conglomeration of bundled, bad social constructs only Fascist Old White Men (FOWN) could have thunk up and used to oppress all-equal primatekind with.

    Shame on them. That the pelvices of Asians are wider than those of Caucasians whose are wider than those of Semites whose are wider than those of Africans whose are wider than those of Bonobos is nothing to do with the higher skull/brain volumes (we are all Africans!), no : this is an athletic feature and therefore real (as opposed to intelligence, or beware, even the idea of measuring that black unicorn of racism).
    Just a thought – there is of course no absolute definition of ‘intelligence (is humanity an ‘intelligent species’ ?) – we measure part of human intelligence in verbal abilities, but that does mean other species are less intelligent than we are if they have other means of communication, or none.
    We would have to adapt our measures to them, to be fair, right ? Chimps are better than us at math.

    So let’s stay with the humans.
    Where would the herd-animal desire for egalitarian “fairness” lead us ?
    Where would fairness lead evolution ?
    Not exactly to positive excesses is my guess.
    DID you know that the genetic differences between Central African and Caucasian Homos are slightly more significant than those between Bonobos and Chimps, which are different species ?

  35. @RaceRealist88

    Let’s put it this way: If you had invested all your money in a software start-up, would you prefer to hire a person who scored 80 on an IQ test as your software developers because “IQ” is arbitrary? Or would you hire the person who scored 140 because you know it measures something real?

  36. eugyppius says:
    @res

    Throughout Cronbach/Meehl belabor that they are only trying to characterize and more explicitly theorize what researchers were already doing. They are not inventing any new demands or methodology. Implicit and at points even explicit is that “construct validation” is what psychometricians are at that moment engaged in and actual examples from the midcentury literature on intelligence tests are adduced to illustrate how construct validation works in the real world.

    TLDR: This paragon of learning has been typing incoherently for months if not years that IQ tests are not “construct valid”, when in fact construct validation applies not to any test or measure but to the procedures and intents employed by researchers in psychology; and the scholars who developed the concept used it to describe procedures and orientations at that moment actually observed among researchers administering intelligence tests. In other words, if anything is construct valid, it is that which RR88 claims is not, very nearly by definition.

  37. Anonymous[299] • Disclaimer says:
    @JoannF

    Excellent points. It’s only a matter of time before we see Affirmative Action Bonobo quotas in Western universities. That would, surely, break the Seventh Seal and herald the coming apocalypse. Bring it on.

    #bonobouniversitylivesmatter

    • Replies: @JoannF
  38. CanSpeccy says: • Website
    @JoannF

    A fun article with a few hilarious reactions.

    Some, perhaps, will find your “reaction” to the reactions of others hilarious. Others may see it as drivel. But let’s get one thing clear: when you say:

    there is of course no absolute definition of ‘intelligence’

    you are gibbering like a bonobo.

    Intelligence is:

    the ability to acquire and apply knowledge and skills,

    including knowledge and skills far beyond anything covered by an IQ test.

    Applied without restriction to humans alone, it follows from that definition that every form of life is intelligent, i.e., capable of acquiring knowledge and using it in an adaptive way.

    The only points at issue, therefore, for those seeking to measure intelligence, are:

    (a) how to measure the capacity for knowledge and skills acquisition, and

    (b) how to measure the the application of knowledge and skills.

    IQ-ists have this simple-minded idea that a few paper and pencil puzzlers is all there is to assessing capacity for the acquisition and deployment of knowledge and skills, whereas, in reality, the task requires the measurement of, in the present state of knowledge, incommensurables.

    For example, no one has the slightest idea how you could compare the abilities of, say, J.S. Bach and James Clark Maxwell, to acquire and use information, or how to compare either of those geniuses with, the world’s most skillful brain surgeon, or finest poet. The IQ-ists don’t even see that as a question they ought to be thinking about, which is why their business looks more like a racket than a science.

    • Replies: @JoannF
    , @Some Guy
  39. Bert says:
    @RaceRealist88

    “A test is valid if and only if there is a theory for what is being measured.
    There is no theory of what is being measured on IQ tests.
    Therefore, IQ tests aren’t valid.”

    Here is the theory of what is being measured: The likelihood of a population creating an industrial/technological civilization is dependent on the population’s ability to investigate and manipulate natural phenomena. Such an ability resides in the subset of the population which can think effectively about abstractions. The size of that subset, as a percentage of the population, is correlated with the average ability for abstract thinking characteristic of the population. IQ tests measure the ability for abstract thinking, and hence the frequency of individuals capable of highly abstract thinking, and hence the likelihood of a given population creating industrial/technological civilization.

    Now I’m sure you won’t enjoy that statement of the theory of what is being measured because empirically the outcome of tests of the capacity for abstract thinking predicts a low likelihood of creation of high civilization outside of Europe and Northeast Asia. Strangely good fit with reality wouldn’t you say.

  40. Yeah, you’re “objective” alright.

    “Objective” means starting with a conclusion and then scrabbling around for “evidence” to support it.

    Good work. You did that well. Lots of graphs means you’re smart.

    • Replies: @segundo
  41. @Hypnotoad666

    you would also be compelled to throw out the entire discipline of psychology as pseudo-science as well

    Suits me just fine – and I’m by no means an ‘IQ-denier’. (See what I did there…)

    A very large proportion of the most widely-cited studies in psychocharlatanry, fail to replicate: significantly less that half of studies in high-value journals.

    So if faced with a published, widely cited ‘research’ conclusion from these fraudsters, Pr(isBullshit) is significantly higher than Pr(isHeads) in a fair coin toss.

    It is the archetype for questionable research practices, low-power research conclusions, and poor statistical method from data, to technique, to presentation.

    Psych is the poster child for the Replication Crisis, and the charlatans in the field have no desire for their grift to be further exposed… they would rather screech about “methodological terrorists” (a phrase actually used by a Princeton charlatan whose work should be retracted in its entirety based on that histrionic bullshit alone).

    • Replies: @res
  42. And, never forget, little Nigerian girls in England are smarter than you, as proved by standardized tests.

    This doesn’t have anything to do with the solipsistic egotism of the promul-gators of IQ Uber Alles, now, would it?

  43. segundo says:
    @obwandiyag

    Perhaps you should wait for the Nigerian Pidgin translation before you make a final judgement.

  44. @JoannF

    Fascist Old White Men (FOWN)

    You missed a trick there – had you added ‘Racist’ into the mix, your acronym could have been “FROWN”… us StalePaleMales are known for wrinkling our brows in response to ascientific nonsense.

    Disclosure: I’m pretty pale, but am infected with waaaaay to much Darkie DNA to satisfy any of the race-obsessed underperformers one reads in threads like this…

    • LOL: mikemikev
    • Replies: @JoannF
  45. SafeNow says:

    Excerpt from a recent California conversation with a high-end medical office, illustrating no IQ test is needed:

    Me: Hello, I am a patient of Dr. Smith. (not real name). I would like to make an appointment,
    please.
    Receptionist: Have you been here before?
    Me: (realizing I had better mirror her language, and, start making generous use of the auditory pause, um, to aid comprehension): Umm…Yes, I have …ummm..been here before.
    Receptionist: What is your name?
    Me: Um, Robert Marshall. (not my real name)
    Receptionist: What is your insurance, Mr. Robert?

    And so on.

  46. Morris39 says:

    Dr. Thompson
    I have a question if you don’t mind answering. I ask b/c I don’t see it discussed and it strikes me that it may be relevant to the understanding of this topic..
    Accept the genetic influence on intelligence as a valid hypothesis. Can you clarify whether the correlations are based on the chromosome gene distribution, some 20 x 10^3 or on the 3 x 10^9 bases (actual working gene molecules)? Is the relationship between the 2 genetic sets empirically known? If so, does it vary over time?

    • Replies: @James Thompson
  47. JoannF says:
    @Kratoklastes

    Yeah I’m actually no longer in the “most witty abbreviations” game, it gets to be boring, as are all of Internet fora, as is human communication as such – but you’re of course correct. How could I have have missed this !
    The only race worth breeding at this point, and after wasting millenia without seeing the obvious, is one of way more intelligent humans – what else could we hope for other than more humaniform intelligence, or be replaced by AI, which is much more likely at this point – and that is all I would be interested in as a serious “racist.” Anti-racism by principle however seems to be in this context a satanic, anti-evolutionary destructivism that endeavors to turn the universe static.

  48. Anonymous[241] • Disclaimer says:

    24 sec on the trail-b test. Beat that suckers!! 🙂

  49. JoannF says:
    @Anonymous

    The world, human society as such, is not yet prepared for this growth – but it’s the obvious way to go.
    You can’t stop progress, and it will come at us from two sides – Bonobos from one and self-programming, exponentially growing AI from the other. will Transbononoism reach (which) human levels, will Transhumanism reach the lower IQ-levels of future AI ? And how do genders fit into all of this ?

  50. JoannF says:
    @CanSpeccy

    “…you are gibbering like a bonobo.”
    I would like to say that this is the most intelligent ad dominam I had lately, but it would be an exaggeration.
    “Intelligence is: the ability to acquire and apply knowledge and skills…”

    In the context of our evolution, current knowledge, in the way that we so far understand things and in the light of our own sad intellectual limits this is more or less the way we defined it.

    I wouldn’t however on the whole want to be limited by that for what’s to come.
    There may be surprises.

    Carry on inside the box – there’s no reason to be ashamed.

  51. utu says:
    @res

    Looking at Table S7 (the first sheet), the numbers you want are there. Now feel free to complain about how small the various R^2 are. Which would at least be a reasonable complaint.

    What values we can see in the Table S7. Polygenic profile score prediction of cognitive performance and educational attainment in Generation Scotland. ?

    R2=3.96% or at most R2=4.74%

    Don’t you think that they are dishonest obfuscating fuckers? That the main body of their paper that includes abstract, discussion and conclusion does not reflect how IRRELEVANT and INCONSEQUENTIAL their work is? That whatever polygenic score they were able to construct using SNPs they found could merely explain no more than 5% of variance in data? Why does this fact does not jump at you when you try to read their paper? They just thought it was not important to include the fact that no more than 5% of variance in cognitive test could be explained with their PGS? Come on res, come on Dr. Thompson, you know perfectly well that the authors are dishonest obfuscating fucker!

    • Replies: @CanSpeccy
  52. CanSpeccy says: • Website
    @utu

    Hey Utu. Love your candor!

    Thing is, IQ-ism is the science of trivial effects. These obfuscating fother-muckers are just doing what IQ-ists has been doing forever.

    Which is sad, really. As a scientific challenge, it would surely be fascinating to investigate the information handling capacity of the brain in a quantitative fashion. For example, after hearing, say, Allegri’s Miserere, I have no clear recollection of what I have heard, yet Mozart is said to have transcribed the whole thing after hearing it just once. From that kind of comparison one could presumably make a quantitative assessment of differences among individuals in relative rates of acquisition of information of particular kinds.

    Processing of information, at least of certain types of processing, might also be quantified. In that way once could develop a real science of intelligence, a science that would undoubtedly confirm what common experience shows, that individuals differ greatly in their profile of intellectual strengths.

    But then we’d be back to reality, with individuals displaying a mish-mash of abilities, particular strengths or weaknesses emerging in unexpected and unpredictable ways.

  53. Austro says:
    @RaceRealist88

    Your reply reads very post modernist. I’m not a scientist but having lived in Sub-saharan Africa most of my life I at least managed to figure out (as a child) that some people just have better problem solving skills than others. Time was irrelevant, as I discovered that no amount of patience would allow the problem to be solved.

  54. buckwheat says:

    Forget all the IQ psycho babble, if you’ve ever hired a black then you already know they are lacking in intelligence. By far the dimmest employees I’ve ever had were black, their IQ, laziness and general stupidity is unequaled by any other race. You don’t need a report to tell you what every already knows………

    • Replies: @Dieter Kief
    , @mikemikev
  55. @szopen

    Just to be Inclusive, there should be tests for “GQ”, [Gansta Quotient] that is those measured traits that are correlated with success in street corner drug dealing, pimping, intimidation, and reflex anger and assault. I mean, let’s be fair and exclude a significant segment of our population for which they have genetically light loading.

    And do not attempt to value “White European” values with Trans African values of asymmetrical transactions, such as extortion, thievery, etc. It’s just another lifestyle, parallel and equal. All things are equal; a belch is just as eloquent as sonnet.

    Well, that’s it. I’m off to my Socialist Intenationale Central Committee Politburo meeting at Beneath Birth Jewish Communistity Center. Buh bye.

  56. @szopen

    “So you are basically arguing that tests are not for intelligence, but only for social class skills; and it’s not intelligence which is genetically based, but traits which help you in modern world to advance to middle class (or to fall below; because IQ tests are different within families). In other words, you don’t believe intelligence is heritable. Instead you surely must believe that class status is heritable (i.e. it’s not that you are born intelligent; it’s just you are born with set of traits which will doom you to fall to lower class, or to raise to high class).”

    Pretty much. Other factors than social class—quite obvously—dictate test scores which is why people in the same family would score differently. Why would class-specific traits be a “rough proxy of intelligence”?

    • Replies: @Aft
  57. @szopen

    “It’s like trying to create tests for health.”

    Who is the healtiest? Who is least healthy?

    The doctor who drinks 6-10 drinks every 24 hours; a college student studying for exams; a teen who was born weighing 2 pounds and shows problems with cognition; or a paraplegic who uses a hand-crank bike to get to work?

    The WHO, in 1946, state that “Health is a concept of complete physical, mental and social well-being and not merely the existence of disease of infirmity.” While Barr (2013) states that “Health involves the health of the body, the health of the mind and the emotions, and the health of the social context in which one lives.”

    You first need to define “health”, then you can worry about constructing tests gauging the “healthiness” of different people. Just like you need to satisfactorily define “intelligence” and then construct a test gauging the “intelligence” of different people.

    “What sense of fair? If it would predict, on average, equally good the outcomes in West-African society for whites, blacks etc (i.e. the result “100” would predict same outcomes for whites, blacks…) then IMO it would be fair.”

    As in, would one group be disadvantaged due to not having the requisite knowledge? The test would be constructed how current tests of ability are constructed—excluding items to fit preconceptions of the test’s designers. Let’s then say that these Kpelle-tests are related to Kpelle-schooling and they predict success in Kpelle-schools. This would then mirror what occurs in America—the thought experiment shows that IQ tests are culture-bound—as all tests of ability ultimately are.

  58. @mikemikev

    “Why do people like you call yourselves the opposite of what you are? Is it a genetic tendency?”

    I believe that our racial categories pick out real kinds in nature. So I’m a race realist.

  59. @eugyppius

    “You misrepresent Jensen. He is talking specifically about items in “tests of general information,” not IQ tests broadly. Jensen’s conclusion is absolutely not that “IQ test scores” are “subjective”, and not even that information (sub)tests are subjective.”

    I do not “misrepresent Jensen” at all—the same definitely is true of IQ tests—which the process of item analysis is.

    “From this we learn three things, all of which you have obscured or misrepresented in a very short space: 1) Raven’s original items corresponded directly to the mental abilities he wanted to measure (i.e., quantitative pairwise progression, figure addition or subtraction, distribution of 2-3 values, etc.). 2) Raven used his “intuition” to rank the difficulty of the items. 3) The relative difficulty of items on modern Ravens matrices tests is established according to “normative data.” One struggles to see what room is left for intuition here. Also you could perhaps read to the end of this article (it is not very long) to encounter some of the theories of intelligence and abstract reasoning that you claim over and over do not exist.”

    The key point there is “without regard to underlying processing theory.” Articulate the theory of intelligence and then relate it to IQ tests and physiological variables.

    “People like Castles, who inter alia uses Gould’s assault on g to (try to) take down Spearman; and people like Richardson, who is so afraid of IQ testing that he wants to ban it, but who also believes IQ tests measure nothing clear and predict mostly results on similar tests”

    What’s wrong with Castles’ argument? They are built to predict results on the similar tests; IQ tests are different versions of the similar tests.

    “Where do large-scale cultural differences come from?”

    Interactions with other humans in a specific niche. Where do they come from? Is your view a reductionist one?

    “Then why are precisely culturally specific (or culturally more specific) items less “g-loaded” (i.e., less subject to correlation) than items that are not?”

    Source? Why are you using “g-loaded” as a signifier (and as if it means anything in a physiological manner) as if it’s not built into the test through construction?

    “Uh, this is not true?”

    Articulate the theory and then show how it fits with modern-day IQ tests—make sure to pick out a test as an example.

    • Replies: @Aft
    , @eugyppius
  60. Some Guy says:
    @CanSpeccy

    For example, no one has the slightest idea how you could compare the abilities of, say, J.S. Bach and James Clark Maxwell, to acquire and use information, or how to compare either of those geniuses with, the world’s most skillful brain surgeon, or finest poet.

    Musician: google “musical aptitude test”.

    Physicist: Quantitative and visual-spatial tests.

    Surgeon: Tests of concentration and dexterity, I guess? Don’t know much about surgeons.

    Poet: Verbal ability, also visual-spatial ability for the kind of poetry that “paints a picture”.

    And of course all those different types of ability tests correlate with each other*, that’s why general IQ tests exist and are useful.

    Then there are of course personality tests that also explain a lot of the variance, various facets of “Openness to experience” for science and the arts for example, while conscientiousness and emotional stability will be particularly important for a surgeon.

    *see for example https://pumpkinperson.com/2015/11/08/do-you-need-to-be-a-genius-to-be-a-genius/

    • Replies: @CanSpeccy
  61. TRM says:

    What is the definition of race? Skin color? Percentage cro-magnon/neanderthal?

    What about quasi-genetic factors, those that change gene expression but not the genes themselves? Especially those that can last for more than a generation.

    I’m thinking specifically of cold water fish intake in the diet. The gene expression effects of EPA/DHA have been shown for 3 generations. What your grandmother ate will affect her, her children and you even if her children and you don’t consume EPA/DHA.

    It can take over a hundred years for the full effects, positive and negative, to become evident.

    I’ve always wondered if the cold water fish intake idea was correct. Iceland, Japan, coastal China, Korea and northern Europe all had (some still have) higher cold water fish intake. Places like Africa, south America, India don’t.

    • Replies: @James Thompson
    , @Aft
  62. Aft says:
    @res

    Not referring to your comments, just the anti-IQ brigade.

    Given the lower sample sizes and number of SNPs found, the low % of variance explained is not surprising in comparison to larger studies like this:
    https://www.nature.com/articles/s41588-018-0147-3

  63. Aft says:
    @RaceRealist88

    You know you’re spouting nonsense but might as well provide links for others:

    https://randomcriticalanalysis.com/2015/11/25/no-the-sat-doesnt-just-measure-income/

    Even controlling for SES, different racial groups see different average scores

    If the SAT truly only measured income, we would also expect the differences across racial groups to be equalized once we control for income. We certainly do not find this. To the contrary, low income whites typically perform about as well as high income blacks (holding income constant we find a bit less than a one standard deviation gap between whites and blacks).

    https://i1.wp.com/randomcriticalanalysis.com/wp-content/uploads/2015/11/google-chrome75.png?zoom=3&resize=350%2C207&ssl=1

    https://i2.wp.com/randomcriticalanalysis.com/wp-content/uploads/2015/11/google-chrome98.png?zoom=3&resize=339%2C361&ssl=1

  64. Aft says:
    @RaceRealist88

    Articulate the theory of intelligence and then relate it to IQ tests and physiological variables

    Everything you are saying is contradicted by Jensen’s work on reaction times: https://lesacreduprintemps19.files.wordpress.com/2012/11/clocking-the-mind-mental-chronometry-and-individual-differences.pdf

    Forget culture, forget tests, IQ shows up in the simplest sub-one-second tasks.

  65. eugyppius says:
    @RaceRealist88

    I do not “misrepresent Jensen” at all—the same definitely is true of IQ tests—which the process of item analysis is.

    It is very hard to know what this word salad means. But even internet trolls must sooner or later read their own citations. Arthur Jensen’s Bias in Mental Testing is a classic. Within its nearly 800 pages you would find a refutation of basically all the shoddy arguments you throw up against the validity of IQ testing. Here is a .pdf:

    http://emilkirkegaard.dk/en/wp-content/uploads/Bias-in-Mental-Testing-Arthur-R.-Jensen.pdf

    Once you have downloaded that, you can go to page 147 of chapter 5 to study Jensen’s subsection on “General Information.” Even this brief discussion is full of ideas that you would do well to internalize. Let us read it together.

    The range of a person’s knowledge is generally a good indication of that individual’s intelligence, and tests of general information in fact correlate highly with other noninformational measures of intelligence. For example, the Information subtest of the Wechsler Adult Intelligence Scale is correlated .75 with the five nonverbal Performance tests among 18- to 19-year-olds.

    So Jensen is distinguishing between informational items and noninformational items on intelligence tests. Imagine that. The two are highly coorrelated, because, as your case illustrates, the ability to learn is governed by native intelligence. Nevertheless, the correlation is not perfect. Mental acumen in and of itself does guarantee that one has received an education, in much the same way as a powerful engine is not by itself a guarantee that a given car has been driven at extreme speeds.
    Jensen goes on:

    Yet information items are the most problematic of all types of test items. The main problems are the choice of items and the psychological rationale for including them. It is practically impossible to decide what would constitute a random sample of knowledge; no “populatoin” of “general information” has been defined. The items must simply emerge arbitrarily from the heads of the test constructors.

    Your favorite sentence is right at the end. You can see quite clearly that the arbitrarily emerging test items are information as opposed to noninformation items.

    No one item measures general information. Each item involves only a specific fact, and one can only hope that some hypothetical general pool of information is tapped by the one or two dozen information items that are included in some intelligence tests.

    Now comes a very interesting and important paragraph:

    Information tests are treated as power tests; time is not an importact factor in administration. […] The items are steeply graded in difficulty. The twenty-nine Information items in the [Wechsler] run from 100 percent passing to 1 percent passing. Yet how can one claim the items to be general information if many of thtem are passed by far fewer than 50 percent of the population? Those items with a low percentage passaing must be quite specialized or esoteric. Inspection of the harder items, in fact, reveals them to involve quite “bookish” and specialized knowledge. The correlation of Information with the total IQ score is likely to be via amount of education, which is correlated with intelligence but is not the cause of it. A college student is more likely to know who wrote The Republic than is a high school dropout. It is mainly because college students, on average, are more intelligent than high school dropouts that this information item gains its correlation with intelligence. The Information subtest of the [Wechsler], in fact, correlates more highly with amount of education than any other subtest.

    Here we read something that others have tried to point out to you in the past, but of which you have remained willfully ignorant. Even culturally biased tests inevitably end up measuring innate intelligence because the two are confounded. Your ability to internalize cultural knowledge is governed by your intelligence. Innately intelligent people tend to learn more and thus to do better on information tests – that is to say, on tests that depend on cultural knowledge. At the same time, noninformation items are an even better measure of IQ than informational items. You will respond like an internet toddler with demands that I source this statement, and like an internet toddler you will not see that I already have, and that the source is the very work you cite, namely Jensen. I invite you to continue reading Bias in Mental Testing, as even a cursory knowledge of its arguments will improve your troll game immeasurably.

    As for the Raven, you write:

    The key point there is “without regard to underlying processing theory.” Articulate the theory of intelligence and then relate it to IQ tests and physiological variables.

    What you do not seem to realize, and this is astounding, is that the very article you’re citing, the article from volume 97 of the Psychological Review published in 1990 by Patricia A. Carpenter, Marcel Adam Just and Peter Shell, “articulate[s]” the “underlying processing theory” that you demand. To glean this much you would only need to read as far as the subtitle, which is “A Theoretical Account of the Processing in the Raven Matrices Test.” When the authors of this paper write that the Raven test was developed without regard to an underlying processing theory, they are not criticizing it, as nobody but charlatans would hold that a “processing theory” must precede the tests that produce the data upon which such a theory would rest. They are pointing out what is already implicit in the design of the test and what they hope to contribute via their own analysis (the processing theory) in subsequent pages.

    What’s wrong with Castles’ argument? They are built to predict results on the similar tests; IQ tests are different versions of the similar tests.

    Castles, who as an advocate for the mentally disabled must have endeared herself to you (see her first book: We’re People First: The Social and Emotional Lives of Individuals with Mental Retardation) does not have an argument. Like many authors of her ilk, she has a few pages on Binet, a few pages on Spearman, and then jumps straight ahead to Gould’s assault on g. You can tell that this is an argument from bad faith by the fact that it is tactical. In this case, the tactic centers upon avoiding an entire century of post-Spearman research and elevating the broadly discredited and dishonest arguments fielded by Gould. When you cite people like this as if they had arguments rather than tactics, you are like a rat claiming to be a fan of the Pied Piper’s music. Either you are a deluded fool, or you’re in on the game.

    “Where do large-scale cultural differences come from?” Interactions with other humans in a specific niche. Where do they come from? Is your view a reductionist one?

    I’m not sure what you can possibly mean by “niche”, but otherwise you appear to believe that cultural differences come from cross-cultural interactions, which is a bizarre theory indeed and also a circular one. It is like saying that abiogenesis is not a problem because clearly living organisms come from other living organisms.

    “Then why are precisely culturally specific (or culturally more specific) items less “g-loaded” (i.e., less subject to correlation) than items that are not?” Source? Why are you using “g-loaded” as a signifier (and as if it means anything in a physiological manner) as if it’s not built into the test through construction?

    I know that many children believe crying “source” is a judo move that defeats all internet opponents. Alas in your case that is not so: If you yourself have in a prior comment in the same thread cited one of the most imposing sources that establishes precisely the thing for which you demand a source (see Jensen, Bias in Mental Testing, passim but especially Chapter 6: Do IQ Tests Really Measure Intelligence?) this tactic is however more akin to punching yourself in the face.

    Articulate the theory and then show how it fits with modern-day IQ tests—make sure to pick out a test as an example.

    Because your demand is incoherent (“fits with”?) it is easy to ignore, and because one such “theory” is in fact outlined and possibly also articulated (your favorite verb!) by your very own paper (Carpenter, Just and Shell: see especially from page 415! but do read to the end), complete with references to “a test as an example”, you make it easy to laugh at you.

    • Replies: @Aft
  66. @mikemikev

    It’s a flashing red warning sign that says “DO NOT FEED THE TROLL”, for those able to read.

    • Replies: @szopen
  67. @TRM

    Have you a reference to the three generation effect?

  68. TRM says:

    8 of them. I first found out about it from a diet researcher named Dr Barry Sears (Zone Diet). At the bottom of this page is the 8 studies he references for that article:

    https://blog.zonediet.com/drsears/blog/diet-can-affect-family-genes-for-generations

    • Replies: @res
    , @James Thompson
    , @Aft
  69. szopen says:
    @Mr. Rational

    Actually, IIRC RaceRealist once had quite different views on IQ and race, and he just changed his mind somewhere on the way; and IMO he was too lazy to change his nickname. That’s quite understandable and I wouldn’t attach to much meaning to this fact.

  70. CanSpeccy says: • Website
    @Some Guy

    And of course all those different types of ability tests correlate with each other

    Yes, they correlate with each other — poorly. Actually, very poorly. And as for correlations between spatial ability, musical ability and the hand-eye coordination of a surgeon, please give us the reference showing any correlation at all.

    Source: Dumb Intelligence Tests: Or Why IQ Testers Need to Improve Their Understanding of Intelligence

    • Replies: @Some Guy
    , @eugyppius
  71. res says:
    @TRM

    Thanks. Here is a link providing full text for the first study (which looks most relevant to me): https://www.researchgate.net/publication/40760124_The_Decrease_of_n-3_Fatty_Acid_Energy_Percentage_in_an_Equicaloric_Diet_Fed_to_B6C3Fe_Mice_for_Three_Generations_Elicits_Obesity

    Abstract
    Feeding mice, over 3 generations, an equicaloric diet in which alpha-linolenic acid, the dietary precursor of n-3 polyunsaturated fatty acids, was substituted by linoleic acid, the dietary precursor of n-6 polyunsaturated fatty acids, significantly increased body weight throughout life when compared with standard diet-fed mice. Adipogenesis observed in the low n-3 fatty acid mice was accompanied by a 6-fold upregulation of stearyl-coenzyme A desaturase 1 (Scd1), whose activity is correlated to plasma triglyceride levels. In total liver lipid and phospholipid extracts, the sum of n-3 fatty acids and the individual longer carbon chain acids, eicosapentaenoic acid (20:5n3), docosapentaenoic acid (22:5n3), and docosahexaenoic acid (22:6n3) were significantly decreased whereas arachidonic acid (20:4n6) was significantly increased. In addition, low n-3 fatty acid-fed mice had liver steatosis, heart, and kidney hypertrophy. Hence, reducing dietary alpha-linolenic acid, from 1.02 energy % to 0.16 energy % combined with raising linoleic acid intake resulted in obesity and had detrimental consequences on organ function.

    Has anyone done research quantifying how quickly and with what magnitude the effect appears/disappears across the generations if the diet is changed in both directions? For example, how does the effect size compare for their three generation case with a case where the most recent generation reversed the dietary change?

    A possible experimental design would be to look at three generations with various possible dietary profiles across the generations. In these terms I understand they looked at 333 and 666 while I would want to see 336, 366, 663, and 633 as well.

    BTW, I consider this omega 3/6 effect one of the more plausible major causes of the obesity epidemic (another being HFCS).

    • Replies: @Aft
    , @Aft
  72. Morris39 says:

    @ Dr, Thompson
    Have you noticed my question, Post #47? I am not making any sort of argument but I am curious if gene expression changes with say age, changes the measured g. If you prefer you can e-mail me.

  73. @Morris39

    The work is done on SNPs, not base pairs.

    We first aimed to determine the genetic contribution of g to the variation in each of the cognitive traits using molecular genetic data. We used a multivariable version of Linkage Disequilibrium Score Regression (LDSC) (25) implemented in Genomic SEM (26) to estimate genetic correlations among the cognitive traits from molecular genetic data. Prior to this formal modelling, we conducted exploratory analyses on the cognitive traits’ genetic correlations, similar to those often conducted on cognitive phenotypes.

    More explanations in the references and supplementary tables.

    As far as I know, age does not affect these scores.

  74. @buckwheat

    You don’t need a report to tell you what every already knows………

    = What everybody you know and/or look upon as reasonable knows. Ok; but then, these people are by far not all people. And it might be for at least some people in the group of those, the difference between the people you think of in your post and the all the others marks, that those reports might turn out to be useful for.

    So: No need to worry. The scientific findings around here in Dr. Thompson’s enlighted blog serve reason quite nicely!

    – Other than that minor quibble of mine, I found your post interesting – thank you!

  75. Some Guy says:
    @CanSpeccy

    “there is about a 0.35 correlation between IQ and physical coordination according to technical studies by the U.S. Department of Labor. ” – https://pumpkinperson.com/2015/11/08/do-you-need-to-be-a-genius-to-be-a-genius/

    Too lazy to find the original study.

    Of course unrelated narrow ability tests correlate poorly with each other, the point is you can use a composite of those tests to create IQ-tests that do correlate highly with things like verbal or quantitative ability.

  76. CanSpeccy says: • Website

    narrow ability tests correlate poorly with each other, the point is you can use a composite of those tests to create IQ-tests that do correlate highly with things like verbal or quantitative abilit

    Actually, the point is that you cannot.

    That’s why places like Harvard and CalTech use separate SAT tests for math and verbal ability. Some people score high on both. Some, like Richard Feynman, do not. And, if you expanded the range of “narrow ability” tests to include a wide range of abilities outside the normal range of psychometric testing you’d find equally striking discrepancies, savants providing the most extreme cases.

    The trivial correlations you see among “narrow ability” tests, i.e., g, could be attributed to variation in factors as unspecific as mitochondrial efficiency, or cardiovascular health.

    And bear in mind that “narrow ability” tests measure abilities subject to major cultural influences. For example, on the Wechsler test, the “similarities” component, which has among the highest g-loadings, is also subject to the largest Flynn effect.

    What that shows is that whatever it is that IQ-tests measure — if they measure anything other than an IQ test scores — is not a purely innate characteristic, but a product of education, culture, etc.

    That’s why international IQ test comparisons tell you nothing about population differences in innate ability.

    • Replies: @Some Guy
  77. Some Guy says:
    @CanSpeccy

    That’s why places like Harvard and CalTech use separate SAT tests for math and verbal ability. Some people score high on both. Some, like Richard Feynman, do not.

    Sigh…

    You started this “debate” by saying that there were no metrics to distinguish between the abilities of physicists and poets etc, and suddenly you remember that SAT tests distinguish between math and verbal. And yes, the two tests correlate highly, 0.68 by the first study I could find, https://www.statcrunch.com/5.0/viewreport.php?reportid=21828) , which is generally considered a high correlation.

    Narrow ability tests = Most useful for predicting success in narrow abilities
    Broad ability tests(verbal, quantitative etc) = Most useful for predicting success in specific fields and subjects
    IQ tests = Most useful for predicting success broadly in a single score

    For example, on the Wechsler test, the “similarities” component, which has among the highest g-loadings, is also subject to the largest Flynn effect.

    What that shows is that whatever it is that IQ-tests measure — if they measure anything other than an IQ test scores — is not a purely innate characteristic, but a product of education, culture, etc.

    IQ tests can’t directly measure the g-factor, so yes that’s true. The g-factor is “innate” though.

    “That’s why international IQ test comparisons tell you nothing about population differences in innate ability.”

    IQ not being a perfect measure of g doesn’t mean it tells you nothing. In fact it appears to correlate remarkably well with innate ability: https://www.researchgate.net/figure/Correlation-between-EDU3-PGS-and-population-IQ_fig1_332076417

  78. CanSpeccy says: • Website

    Sigh…

    Sigh… What does that mean? Hey thickie?

    Hardly evidence of a rational protagonist. Rather evidence of an intellectual bully.

    You started this “debate” by saying that there were no metrics to distinguish between the abilities of physicists and poets etc, and suddenly you remember that SAT tests distinguish between math and verbal.

    Sigh…

    How does the existence of metrics to distinguish between certain number-related skills and certain word-related skills, prove that there are metrics to distinguish between physicists and poets. Or do you really think you can pick out a future Shakespeare with a SAT test?

    Narrow ability tests = Most useful for predicting success in narrow abilities
    Broad ability tests(verbal, quantitative etc) = Most useful for predicting success in specific fields and subjects
    IQ tests = Most useful for predicting success broadly in a single score

    Except that as noted here, the usefulness of psychometric tests in predicting success is rather slight. Furthermore, an inaccurate test of future potential may do harm not good, especially when sold by people insisting, as you do, that test performance is destiny.

    True, bullshit estimates of future potential may not hold everyone back. Beverley McLachlin, for example, Chief Justice of the Canadian Supreme Court, whose Grade 8 teacher informed her that she had “no useful skills.” For some, I suppose, such predictions serve as a spur to greater effort. Still, quackery is quackery.

    IQ tests = Most useful for predicting success broadly in a single score

    Yeah, like Richard Feynman’s score of 123! That really had him pegged. Likewise, the exclusion of Louis Alvarez and William Shockley from the Terman study of high IQ individuals.

    Sigh… indeed.

    • Replies: @Some Guy
    , @eugyppius
  79. Some Guy says:
    @CanSpeccy

    Sighs just means I’m tired of the debate, not an insult.

    How does the existence of metrics to distinguish between certain number-related skills and certain word-related skills, prove that there are metrics to distinguish between physicists and poets. Or do you really think you can pick out a future Shakespeare with a SAT test?

    What you asked for was metrics that can distinguish the abilities of a physicist and a poet, not to predict with 100% accuracy who would become a poet. Physicists will be relatively stronger on quantitative ability and poets relatively stronger on verbal ability. There, they’re distinguished.

    Except that as noted here, the usefulness of psychometric tests in predicting success is rather slight. Furthermore, an inaccurate test of future potential may do harm not good, especially when sold by people insisting, as you do, that test performance is destiny.

    When did I do that? Obviously the weight attributed to psychometric tests must be proportional to their predictive validity, anything else is misuse.

    Beverley McLachlin, for example, Chief Justice of the Canadian Supreme Court, whose Grade 8 teacher informed her that she had “no useful skills.”

    So, skimming that link, the test accurately assessed her excellent reading-retention, but the teacher dismissed it because she was a girl? That’s clearly a point in favor of tests and a point against subjective teacher assessments.

    Yeah, like Richard Feynman’s score of 123! That really had him pegged. Likewise, the exclusion of Louis Alvarez and William Shockley from the Terman study of high IQ individuals.

    I believe it was 125, which is higher than 95% of the population. His test result clearly hinted at his potential. As I said, the weight attributed to psychometric tests must be proportional to their predictive validity. No one has ever claimed that one IQ test is perfectly correlated to success, especially in childhood. A factor can be important without being everything, why is that so hard for people to understand? Like how you’re built is important but not everything when it comes to sports.

    I’ll also note that all those examples are from when IQ tests were quite new and emphasized verbal ability.

    • Replies: @CanSpeccy
    , @Aft
  80. mikemikev says:
    @buckwheat

    I know right. Imagine requiring a complex mathematical proof to accept that Africans are a bit dim.

  81. eugyppius says:
    @CanSpeccy

    Your matrix has been clipped from Carroll, The Higher-stratum Structure of Cognitive Abilities, in: A Scientific Study of General Intelligence, ed. Helmuth Nyborg; see page 15.

    Carroll extracts from the WJ-R subtests strong statistical evidence for g .

    The claim is not that strong correlations prevail over all tests, but that those tests which draw more heavily on g will be more strongly correlated to one another (and that their correlations will reflect the extent to which they are g-loaded).

    Not all of the subtests on the Woodcock-Johnson battery are heavily loaded on g. They are not even intended to be. See Woodcock himself on the cognitive subtests of the WJ-R and their theoretical underpinnings:

    http://www.iapsych.com/wj3ewok/LinkedDocuments/woodcock1990.pdf

    The potted version is that the battery was conceived to test fluid and crystallized abilities (Gf-Gc theory) and the various sub-abilities posited by this theory. Woodcock himself was agnostic about g: on the very last page he says maybe it’s there and says Carroll is crunching a lot of numbers.

    Jump ahead 13 years to 2003 and we have Carroll using WJ-R norming sample data (your matrix!) to find strong evidence of g and only very faint support for the old Gf-Gc theory.

  82. @res

    The problem is what happens once evidence emerges that IQ and economic outcomes are (weakly) correlated: the most passionate on either side of the debate dig their trenches, and armed with bad data devoid of context, start yelling at the opposition.

    I have gotten used to reading dismissals of the lack of Sub-Saharan African economic development as being largely explicable by some gap in cognition – that white Western society advanced dramatically because YT is just smarter, and that the jigaboos just need to shake off 500 years of colonialism and #learntocode.

    Those are really very boring things to read because nobody considers an interesting phenomenon that goes precisely the other way… it is the veritable Dog That Did Not Bark.

    Viz.,

    If Ashkenazim are ~σ smarter than the human herd, how come the Industrial Revolution didn’t begin in the Jewish Oblasts of Eastern Europe? The Pale of Settlement must’ve been chockful of smarties, inbreeding themselves to a Lithuanian Silicon Valley… yeah…, not so much: by the middle of the 20th century that region was 20 years more backward than Appalachia.

    In their natural environment, their lives were benighted; their economic development stunted relative to their countrymen; their life expectancies shorter (and relative to a fairly low base).

    How come that?

    They underperformed their fellow citizens (again, on average).

    Let’s accept arguendo that the claimed ~σ exists, and is not as overstated as the death toll at Bethar, or Jussie Smollettberg’s police complaint. (0.3σ is what you arrive at if you examine the data a bit more critically, as VoxDay will point out to anyone who asks – but let’s leave that aside).

    How come they contributed so little to human advancement (even to their own lives) in the period before the last two or three generations – coincident with massive westward migration to a modern, Western, industrial/technological environment?.

    It’s almost as if – and bear with me here – an arbitrarily selected smart Ashkenazi had just as little chance of advancement in Tsarist Russian (and later in the USSR) as an arbitrarily selected smart black kid has in the modern US.

    It’s almost as ifenvironment matters – and what matters is the specifics, not the general environment (Tsarist Russia was slightly behind most of Western Europe, but not that far; but they treated rural Ashkenazim badly enough to put paid to any difference in mean IQ).

    I don’t really care that much either way, but I just dislike people on both sides who pretend that their schtick has any actual basis in the scientific method (which is man’s greatest gift to man), properly conducted.

    Smart people do better is almost as tautological as ‘survival of the fittest’. It also suits me, because I was lucky enough to have the genetics and upbringing that gave me every opportunity to make the most of such gifts as I had. (By my own estimates I’ve wasted at least as much of that potential as I fulfilled: square pegs and all that).

    Lastly: consider what it means if being smart is entirely (or mostly) genetic. If that’s the case, smart people really don’t get the credit for their achievements – as a very famous Australian football coach said (I’m paraphrasing)

    “Give me a hard worker over natural talent any day: hard work’s your own – you get natural talent from the eye of your father’s cock”

    Actually this can be ‘lastly’: my guess is that society is moved forward by 120-IQ guys who work their asses off. Any more than 1σ higher than that, economic performance should be expected to tail off, because people become less acquisitive and view ‘normal’ workaday stuff as unbearable.

    (To me the old trope about the 140IQ neckbeard living in his Mom’s basement playing Minecraft and covered in Cheeto dust makes perfect sense: satisficing).

    • Agree: utu
  83. eugyppius says:
    @CanSpeccy

    Intelligence as measured by IQ tests predicts socio-economic status strongly: slightly more strongly than parental SES and grades in school. Because intelligence is highly heritable and grades themselves are correlated with IQ scores, all three measures are of course confounded:

    https://tinyurl.com/y4x4nf5z

    Feynman’s IQ of 125 was self-reported and was maybe a joke, but actually the number is in no way implausible. Consider our data on the IQ of academics in different fields:

    http://www.religjournal.com/pdf/ijrr10001.pdf

    The sample sizes are small (not a lot of work has been done on this) but a 1967 survey of physicists at Cambridge yields an average of 127 on the Wechsler. That’s at the higher end of where the professoriate seems to sit. Professors, even very prominent ones, are not massively intelligent (just as they don’t have especially high SES).

    • Replies: @CanSpeccy
  84. CanSpeccy says: • Website
    @Some Guy

    What you asked for was metrics that can distinguish the abilities of a physicist and a poet, not to predict with 100% accuracy who would become a poet. Physicists will be relatively stronger on quantitative ability and poets relatively stronger on verbal ability.

    So there we have it. Human intelligence defined: quantitative ability and verbal ability. Nothing else? No, that’s it. LOL.

    Poetry is not the product of SAT verbal ability. It is about rhyme and rhythm and resonance, alliteration and allusion. An IQ test doesn’t touch on any of those aspects of language.

    Psychometrics will never be a real science until it formulates an operational definition of intelligence. You can’t say we have a test and what our test measures is intelligence. That is like an ichthyologist using a net with a six-inch mesh, arguing that what his net does not catch is not a fish.* But in the case of the IQ-ist, he’s using a net with a yard-wide mesh, and saying what his net doesn’t catch is not intelligence.

    * With acknowledgements to Arthur Eddington for the metaphor.

    • Replies: @Some Guy
  85. CanSpeccy says:

    Intelligence as measured by IQ tests…

    Love that. What has to be established (i.e., that IQ tests measure intelligence) simply assumed. As for

    predicts socio-economic status strongly.

    It doesn’t. According to your own reference, Strenze, Table 1, the correlation coefficients for all studies between income and IQ average 0.21, meaning that, IQ accounts for 4% of variation in income. Pfui!

    As I said, IQ-ism is the science of trivial effects.

    • Replies: @res
    , @eugyppius
  86. CanSpeccy says: • Website
    @eugyppius

    Professors, even very prominent ones, are not massively intelligent (just as they don’t have especially high SES).

    Agreed, most professors are not too bright and do no work of any great intellectual importance either.

    But the question at issue is how come that among the geniuses of 20th century physics, the three who’s IQ’s were measured, did not have extraordinarily high IQs.

    The only conclusion to be draw is that an exceptionally high IQ is not a prerequisite of genius — which is not the same as saying that genius does not require exceptional mental ability. It just means that IQ tests don’t pick up on whatever it is that leads to exceptional intellectual achievement.

    • Replies: @res
  87. res says:
    @CanSpeccy

    As I said, IQ-ism is the science of trivial effects.

    When interpreting this graphic be sure to remember that the lowest scoring group is still within the top 1%.

    • Replies: @CanSpeccy
  88. res says:
    @CanSpeccy

    But the question at issue is how come that among the geniuses of 20th century physics, the three who’s IQ’s were measured, did not have extraordinarily high IQs.

    They are not the only ones with measured IQs. And some of the others have high range test scores available if anyone takes the time to look at them once the data is public.

    https://search.amphilsoc.org/collections/view?docId=ead/Mss.B.R621-ead.xml

    The book version of Anne Roe’s work is available at
    https://www.amazon.com/Making-Scientist-Anne-Roe/dp/0837171512

    Note that one of those scientists is Luis Alvaraz. It would be interesting to see how his test scores there compare with the Terman results you are so fond of citing.

    (Dr. Thompson, any idea if anyone in ISIR would have any interest in following up on the now public material there?)

    P.S. And who’s instead of whose? You’re better than that.

    • Replies: @James Thompson
  89. Aft says:
    @Bardon Kaldian

    Psychometry is great for measuring one’s abilities in various areas (numbers, space orientation, technical skills, words & vocabulary,..), but it cannot measure strong creativity, focus, will, work ethic, mental illnesses that hamper one’s life etc.

    Actually it can. There are tests for literally every single one of those things that explain a decent proportion of the variance in each.

    I’m sure if someone really wanted to they could build a pretty decent composite score of the whole thing: IQ, conscientiousness, work ethic or will, mental illness risks, physical illness risk, etc.

    • Replies: @CanSpeccy
  90. Aft says:
    @TRM

    Largely irrelevant.

    The Chinese even when at the edge of Malthusian starvation for centuries were vastly more intelligent than the groups you mention.

    And today many still contend with fluroide poisoning in many regions, toxic air, a very high arsenic diet (rice has its issues) and still match the developed world in international assessments.

    Nutrients (prenatal, not transgenerational…) matter, but they won’t turn South Africa into Singapore.

    • Replies: @TRM
  91. CanSpeccy says: • Website
    @res

    Hey Res,

    That’s an interesting graph. I mean, talk about torturing the data!

    Still, yes, if you’re good at math at 13 you’re more likely to have some success in a STEM-related career.

    But SAT tests are not the same as IQ tests. They test the ability to to recall and use learned information, whereas IQ tests are supposedly impossible to prepare for (although there’s the issue of test sophistication so the distinction cannot be quite so stark).

    And here’s the thing, the closer you come to traditional academic assessment, the more predictive of actual intelligence are the results likely to be.

    Intelligence is, by common usage, defined as the ability to acquire and use information. But information comes in different streams, visual, auditory, tactile, proprioceptive, etc., each stream processed by a different part of the brain.

    So if you are going to assess intelligence you’ll need to assess the assimilation and utilization of information acquired via multiple channels and processed by different neurological modules.

    That was the sole method of assessment by schools, until they became corrupted by psychologists bearing IQ tests. By the old methods, Richard Feynman was a unique mathematical genius. By an IQ test, he was just bright normal.

    • Replies: @res
  92. CanSpeccy says: • Website

    P.S. And who’s instead of whose? You’re better than that.

    LOL

    Whose is the possessive form of the pronoun who, while who’s is a contraction of the words who is or who has. However, many people still find whose and who’s particularly confusing because, in English, an apostrophe followed by an s usually indicates the possessive form of a word.

    When I wrote who’s, or was it whose, don’t recall now, I had a twinge of anxiety. But after spending decades editing over ten million words of scientific prose, and creating the world’s most intensively cited journal—out of more than 100—in its [or could it be it’s] field, I have little taste for additional proof-reading.

    I don’t know whether it was the way I was taught to read and write or being left handed, or mere stupidity, but there are certain rules and conventions of English spelling that I seem unable to reliably assimilate. With Google to hand, I often check, but sometimes I just wing it.

  93. CanSpeccy says: • Website
    @Aft

    I’m sure if someone really wanted to they could build a pretty decent composite score of the whole thing

    You could be right, but what would you have in the end? Something like a Grade 12 transcript and a perceptive teacher’s character assessment. In other words, IQ testing is a step, no a huge stride, away from accurate assessment.

    OK, if you have to send ten million men and women off in a hurry to fight the Chinese or something, and you need to sort out who’ll give the orders and who’ll be responsible for the latrines, an IQ test may be a useful short-cut. But using IQ tests when time allows for vastly better methods is absurd.

    • Replies: @Aft
    , @res
  94. Aft says:
    @eugyppius

    Masterful.

    (Even though RR88 is a troll, an imbecile, or both, good rebuttabls are fun to read. Also that book is surprisingly broader than I expected from the title, may actually be his best.)

    • Replies: @eugyppius
  95. Aft says:
    @TRM

    That article plays very fast and loose with all this.

    Those who have been constantly exposed to dairy products after weaning have developed an epigenetic programming that seems to be permanent.

    This is very well established to be genetic, not epigenetic.

    Which of the studies provide actual evidence of transgenerational epigenetic changes? Looking at several of them, none of them actually show this.

    • Replies: @res
  96. Aft says:
    @res

    That study is poorly written and poorly documented (with no evidence presented whatsoever of transgenerational epigenetic changes, just that they happened to feed them a horrible diet for a few generations and then only look at the final one).

    • Replies: @res
  97. Aft says:
    @Some Guy

    1. Feynman’s mathematical IQ can be pretty well inferred from his crushing the Putnam exam with basically no preparation: https://mks.mff.cuni.cz/kalva/putnam/putn39.html

    2. Shockley largely stole credit for the transistor from coworkers who’d already figured it out. He added some immaterial tweak to claim it was his invention. His subsequent foray into founding his own company saw him smooth-talk a bunch of geniuses into joining, who then all quit at once and most of their companies succeeded wildly (while his floundered). Some people get ahead from confidence and stealing, not ability.

    Also spatial intelligence, crucial to engineering and design, varies widely and is still poorly captured by many IQ batteries (it doesn’t help its popularity that there is an enormous sex gap on mental rotation; plus Jews have unremarkable spatial intelligence but East Asians may be higher on it than even their mathematical ability).

    Subsequent gifted studies have validated spatial ability tests in predicting exceptional achievement in engineering.

    3. More broadly, of course there are exceptions who had a bad day, weird test form, or had lopsided ability tilt. Roe’s study of eminent scientists is far more accurate and designed to be representive: he found the most eminent scientists averaged around an IQ of 160, four standard deviations above the mean.

    • Replies: @res
  98. eugyppius says:
    @CanSpeccy

    Two things:

    1) For Strenze, SES is income + education + occupation. Not income alone. Much higher correlations with the other two.

    2) The weak correlation with income is obviously interesting.

    It is intuitive: My brother-in-law, for example, is a total idiot, and he earns a substantial salary.

    Of all the factors considered by the Strenze meta-analysis, however (parental income, parental education, father’s occupation, ‘SES index’, mother’s occupation etc.), IQ has the strongest correlation with income, even though it is weak.

    It is easy to complain about small correlations. But ‘small’ risks being a comparative term. Small compared to what? What other factor helps us explain income better?

    Anyway Strenze’s SES claims stand.

  99. @res

    Thank you for reminding me about Anee Roe’s fascinating work. I think you had mentioned it and that it was discussed at ISIR2019. I will see if anyone is interested.

  100. eugyppius says:
    @Aft

    Ha, thanks.
    Yeah, Bias is really good. Clocking the Mind though also cuts through much garbage.

  101. TRM says:
    @Aft

    Centuries? We’ve only had IQ measurements for one century and even that has been subject to constant revision as new information comes forward. Everything before that is way too subjective.

    I agree that prenatal nutrition is crucial but I disagree with you on the transgenerational aspect. The research, so far, does not back your position.

    With a lag time of over 100 years to see the full effects, both positive and negative, it is difficult to study but until we try it you don’t know that South Africa can’t be turned into Singapore. That is just your opinion.

    As far as I can tell culture is the largest changeable factor but even in groups where you have the same cultural valuing of education there are differences. The Han Chinese in the interior are as close as you can get geneticly to the Han Chinese on the coast yet statistically significant differences exist.

    • Replies: @Aft
  102. Some Guy says:
    @CanSpeccy

    So there we have it. Human intelligence defined: quantitative ability and verbal ability. Nothing else? No, that’s it. LOL.

    Again, what you asked for was a test that distinguished between the abilities of physicists and poets, so why would I list all aspects of mental abilities?

    Poetry is not the product of SAT verbal ability. It is about rhyme and rhythm and resonance, alliteration and allusion. An IQ test doesn’t touch on any of those aspects of language.

    So the great poets will have average scores on the SAT verbal then? No, I didn’t think so.

    In math, “a product is the result of multiplying, or an expression that identifies factors to be multiplied.” Verbal ability is obviously a factor in poetry, as is openness to experience etc. You didn’t ask for tests that can predict who is going to be a poet with 100% certainty and I never claimed they could, so stop pretending that’s what we’re debating. Try understanding what “factor” and “variance explained” means, please.

    Anyway, not going to keep debating stupid stuff, have a good day.

    • Replies: @CanSpeccy
    , @Aft
  103. CanSpeccy says: • Website
    @Some Guy

    Anyway, not going to keep debating stupid stuff

    Sigh…

  104. res says:
    @CanSpeccy

    I mean, talk about torturing the data!

    How so?

    • Replies: @CanSpeccy
  105. res says:
    @Aft

    Definitely flawed (and I agree the “proof” that the effect they are observing is caused by epigenetics is lacking), but I do think there is something going on there which is worth understanding better. Especially given that the US seems to be conducting a natural experiment with what happens if we feed multiple generations of humans in a row with a diet having an omega 6:3 imbalance. Problem is, there are so many other variables present it is hard to interpret the results of the natural experiment.

    • Replies: @Aft
  106. res says:
    @Aft

    Feynman’s mathematical IQ can be pretty well inferred from his crushing the Putnam exam

    Well, if you mean “really, really high” ; ), but I hesitate to assign a number.

    Subsequent gifted studies have validated spatial ability tests in predicting exceptional achievement in engineering.

    Excellent point. Two of the other plots from the same source as the one I gave above cover this:
    https://infoproc.blogspot.com/2016/01/smpy-at-50-research-associate-position.html

    I think this one is the best. Arrow length and direction represent spatial ability.

    Law in particular intrigues me. I wonder how much of that is caused by Jewish overrepresentation in the legal profession?

    he found the most eminent scientists averaged around an IQ of 160

    FWIW, Anne Roe is female.

    • Replies: @Aft
    , @James Thompson
  107. Aft says:
    @res

    Way, way north of 160.

    Law tends to be a fallback for people that didn’t affirmatively opt into or stand out in other fields (at least in the States where application comes after undergraduate education and requires the least specific preparation of any graduate school, just a single test and no evidence of any particular interest in law).

    The spatially-oriented aspiring surgeons, engineers, scientists, etc. have already been removed from the pool, so there is some adverse selection there on spatial ability. But not as much as you’d think on math ability because analytical and logical reasoning (especially the LSAT “logic games” section) are still dependent on it to some degree.

    And the Jewish skew certainly impacts those numbers as well.

  108. Aft says:
    @res

    There’s no doubt Darwin was right about pangeneis (shown from studies of sperm methylation in humans already). And that some endocrine disruptors or major toxins can have transgenerational effects.

    There’s also some evidence for body size effects and possibly stress or risk-taking phenotypes. And these make some sense (adapting to presence or absence of resources, being more likely to venture out from highly stressful conditions).

    But these don’t explain race-IQ gaps:

    https://i1.wp.com/randomcriticalanalysis.com/wp-content/uploads/2015/11/google-chrome75.png?zoom=3&resize=350%2C207&ssl=1

    https://i2.wp.com/randomcriticalanalysis.com/wp-content/uploads/2015/11/google-chrome98.png?zoom=3&resize=339%2C361&ssl=1

    it is difficult to study but until we try it you don’t know that South Africa can’t be turned into Singapore. That is just your opinion.

    Good luck with that.

    • Replies: @Aft
  109. Aft says:
    @Some Guy

    The great poets were no dullards:

    https://www.theatlantic.com/magazine/archive/1991/10/the-case-for-oxford/306478/

    Goethe was a precocious child, richly endowed physically and mentally. He absorbed knowledge spontaneously and without effort. His fancy, too, was active, and he took delight in relating the most marvelous tales, which he himself invented, to a company of admiring friends. The two fairy tales, “The New Paris” and “The New Melusine,” which he reprinted in a somewhat improved shape in his autobiography, belong to this period.

    A charming anecdote is related of his fondness for Klopstock’s biblical epic, “The Messiah,” before he had yet emerged from the nursery. Frau Aja, his mother, had surreptitiously borrowed this book, and went about with it in her pocket, because her husband highly disapproved of Klopstock’s wild and rebellious rhapsodies. Goethe and his younger sister Cornelia, sharing their mother’s predilections, therefore committed the precious verses to memory, and amused themselves with personating the enraged Satan and his subordinate fiends. Standing on chairs in the nursery they would hurl the most delightfully polysyllabic maledictions at each other. One Saturday evening, while their father was receiving a professional visit from his barber, the two children (who were always hushed and subdued in his presence) were seated behind the stove whispering sonorous curses in each other’s ears. Cornelia, however, carried away by the impetus of her inspiration, forgot her father’s presence, and spoke with increasing violence:

    “Help me! help! I implore thee, and if thou demand’st it
    Worship thee, outcast! Thou monster and black malefactor!
    Help me! I suffer the torments of death, the eternal avenger!” etc.
    The barber, frightened out of his wits by such extraordinary language, poured the soap-lather over the counsellor’s bosom.

    Johann Wolfgang von Goethe 1749 – 1832 Germany
    • Natural genius – knew Greek, Latin, French, and Italian by age eight; studied law at the universities of Leipzig and Strasbourg; also studied music, art, biology, history, philosophy, chemistry, anatomy, and geology. Pioneered theories about light and color. Developed morphological approach in botany and zoology

    These two would smash through the ceilings of any and all verbal assessments one could possibly devise. Certainly not sufficient, but absolutely necessary.

  110. Aft says:
    @TRM

    Everything before that is way too subjective.

    https://en.m.wikipedia.org/wiki/Timeline_of_historic_inventions

    https://en.m.wikipedia.org/wiki/List_of_Chinese_inventions

    Not found: any wikipedia page on African inventions….

    • Replies: @Aft
  111. Aft says:
    @CanSpeccy

    Of course!

    SATs are correlated north of 0.8 with IQ tests, and for most purposes you’d care more about the SAT with its implicit light conscientiousness loading (or ability to convert gf into gc over time) than raw “IQ”

    https://randomcriticalanalysis.com/2015/06/18/on-sat-act-iq-and-other-psychometric-test-correlations/

    And with the caveat that psychometric (including personality) tests are way more precise and accurate than most teachers opinions or cherry-picked references, e.g.

    https://www.y2cp.com/ressources/publications/articles/tests/wave/how_valid_is_your_questionnaire.pdf

    • Replies: @res
  112. Aft says:
    @Aft

    https://youtu.be/Q-12_M2XDxY (be careful, you can never un-see this)

    Did you have an estimated date for this SA-SG convergence?

  113. Aft says:
    @Aft

    But c.f.

    "We (africans) are the moust intelligent race on the face of the planet…we invented the cars, the computers.."😲 pic.twitter.com/X35BfkCgnv— Nova Ordem Social (@Mariomachado14W) September 27, 2019

    https://mobile.twitter.com/Mariomachado14W/status/1177522957599617024

  114. res says:
    @Aft

    That first link is interesting. Thanks.

    The comments there led me to look at some of Thomas Coyle’s work. Dr. Thompson has covered some of it in his blog. For example: http://www.unz.com/jthompson/tilting-at-racial-difference_7
    But I don’t see anything about Coyle’s work on test residuals after removing g (the topic of the comment at that link which started me off).

    Here are two of his papers on that.
    SAT and ACT predict college GPA after removing g (2008, available on libgen): https://www.sciencedirect.com/science/article/abs/pii/S0160289608000603
    Predictive Validity of Non-g Residuals of Tests: More Than g (2014): https://pdfs.semanticscholar.org/ea73/72291cb9442f7543261a89a5fff6db9c0a61.pdf

    Does anyone have any thoughts on this? (Aft, I get that your “light conscientiousness loading” comment is probably about this)

    • Replies: @Aft
  115. @res

    As you note, the distinction of being female is likely of some worth to Anne Roe.

  116. res says:
    @CanSpeccy

    You could be right, but what would you have in the end? Something like a Grade 12 transcript and a perceptive teacher’s character assessment. In other words, IQ testing is a step, no a huge stride, away from accurate assessment.

    I think you underestimate the value of a measure which reduces the subjectiveness caused by different schools and different teachers. The value is you can get a decent comparison of someone who has had the best primary and secondary education and someone who has never even encountered a perceptive teacher.

    And to be clear, IMO scarcity of perceptive teachers (and difficulty of third parties telling who is and is not perceptive) limits your approach.

    I think the combination of greater objectivity and universality do make IQ (and/or ability) testing a huge stride forward from the regime you describe. But for best decisions I want all three when possible.

    OK, if you have to send ten million men and women off in a hurry to fight the Chinese or something, and you need to sort out who’ll give the orders and who’ll be responsible for the latrines, an IQ test may be a useful short-cut.

    Yes. This seems like a bit of a concession from you (or at least your historical persona here).

    But using IQ tests when time allows for vastly better methods is absurd.

    Perhaps, but resources are almost always limited. One issue is which approaches make for the best pre-screens. As an example, consider graduate school admissions. How would you rate the following for utility (remember, you can’t interview everyone) as pre-screens?
    – Undergrad school attended.
    – GPA
    – In major GPA
    – GRE scores
    – Others of your choice?
    Also, grade inflation has seriously impacted the value of GPA as an indicator.

    Another issue is which predictors add incremental predictive value. If you are ignoring test scores I think you better be sure to try to find another decent proxy for intelligence.

  117. res says:
    @Aft

    This is very well established to be genetic, not epigenetic.

    Could you elaborate? My understanding is that lactose tolerance in adulthood is genetic based, but it is necessary to be exposed to dairy products after weaning for the appropriate enzymes (are any relevant other than lactase?) to stay active. What do you see as the mechanism which separates potential from actual lactose tolerance?

    • Replies: @Aft
  118. Aft says:
    @res

    Original link was pure pseudoscientific gibberish:

    The next question is how long does this epigenetic programming have to be continued until it becomes a permanent part of the gene structure. One indication might be found in the development of lactose intolerance in those populations who have been exposed to dairy products for thousands of years. Seventy percent of the world’s population can’t digest these dietary products because they have lost the ability to maintain the necessary enzyme production after weaning from mother’s breast milk. Those who have been constantly exposed to dairy products after weaning have developed an epigenetic programming that seems to be permanent.

    This is 100% due to a genetic mutation that appeared (maybe once, maybe more than once, haven’t read too much on this) and spread through genetic evolution not “transgenerational epigenetic inheritance”

    • Replies: @res
  119. Factorize says:

    I am working on an assignment about race and ethnicity and I am trying to find an example of a genetic variant that would lead to social fractionation of primary school children along such a boundary. I noticed that this happened in my personal life experience, though it would be helpful if I could point to a specific variant that clearly differed by different populations (especially for those of mostly European ancestry). The assignment and course materials imply that noticing and responding as expected to such differences is racist. However, having a counter-example which demonstrated that this is actually typical human behavior would leave the unsettling implicit message for those who wanted to argue this further that even young children are inherently racist (contra tabula rasa). A genetic example that was less inflammatory than IQ would be best, perhaps pro/anti-sociality?

    • Replies: @Aft
  120. Aft says:
    @res

    Certainly some mix of conscientiousness and interest in written school work.

    Those tests (SAT/ACT) tend to measure some level of life-long investment in reading and in paying attention in math, continent on having high g in the first place. Not surprising they incrementally predict grades after controlling for g.

    I know a fair number of people from my childhood gifted classes who didn’t exactly focus on academics or schooling for too much longer but still have quite a very sharp presence and clearly applied their g to become very verbally (or one might say psychosocially) gifted instead. The colleges probably have it about right if they’re looking for people to go along with their model.

    [And e.g. the MCAT has a massive non-g loading on sitting there and studying massive amounts of biology and chemistry, the LSAT is almost certainly more purely g-loaded than the SAT and ACT, while the GRE used to reward some pretty obscure and pretentious vocabulary. Each test has its own little non-g tilt, even if it’s hard to imagine any of those get a much lower than a ~0.8 correlation with g.]

    Any criticism some of the IQ deniers have about “culture” and “education” apply to the SAT, though certainly not to an actual IQ test.

    A later open-access paper: https://www.mdpi.com/2079-3200/6/3/43/htm

    • Replies: @res
  121. Aft says:
    @Factorize

    No idea of the validity but this is amusing on many levels: https://www.youtube.com/watch?v=Ls-DrVqDM90

    Video I was actually looking for: https://www.youtube.com/watch?v=eHeSlMui-2k (after a couple hundred hours on flights and in airports in Asia, I started to wonder “how are all these children so quiet and well-behaved, what is going on over here?”; seeing this video helped explain a lot of that mystery…)

    And of course the black babies were always running around and the one American Indian baby I saw just kept shouting at every single passerby for hours, etc.

    • Replies: @Factorize
  122. res says:
    @Aft

    In this excerpt:

    Those who have been constantly exposed to dairy products after weaning have developed an epigenetic programming that seems to be permanent.

    I think the exposure enabling continued lactose tolerance is on target (but see below for clarification), but the epigenetic part is speculation. Here is a paper discussing a different mechanism–which I find more plausible.

    https://nutritionj.biomedcentral.com/articles/10.1186/1475-2891-12-160

    As early as 1993, adaptation of the colon bacteria by increasing the exposure to lactose was a suggested approach by Briet et al [8] to improve lactose digestion and tolerance. In 1996, Hertzler and Savaiano demonstrated significant improvement in lactose digestion and tolerance and elevation of fecal beta-galactosidase due to colonic adaptation [9].

    Here is a popular source (not sure how reliable, but sounds plausible) which talks a bit more about this.
    https://www.wellandgood.com/good-food/quitting-dairy-lactose-intolerant/slide/2/

    If you’re wondering how it is that more than a quarter of the population can still tolerate a latte—despite being maldigesters— there’s an explanation. Our gut bacteria actually produce lactase for us. And the more dairy we give them, the more lactase they produce.

    “The bacteria in our colon need to be fed in order to survive,” explains Savaiano. “So whatever you feed them, those bacteria are going to prosper. Individuals who are used to eating lactose in their diet have more lactase enzyme [than people don’t eat lactose-containing foods]—we think six to eight times more—and are more efficient at digesting it so they don’t get symptoms.”

    This recent review paper looks like a thorough look at lactase variation with exposure: Lactose digestion in humans: intestinal lactase appears to be constitutive whereas the colonic microbiome is adaptable
    https://academic.oup.com/ajcn/article/110/2/273/5512720

    From the abstract:

    In conclusion, endogenous lactase expression does not depend on the presence of dietary lactose, but in susceptible individuals, dietary lactose might improve intolerance symptoms via colonic adaptation. For these individuals, lactose withdrawal results in the loss of colonic adaptation, which might lower the threshold for intolerance symptoms if lactose is reintroduced into the diet.

    So it looks like I was wrong in thinking that lactose exposure was necessary for enabling continued lactose tolerance in those who are genetically tolerant (endogenous lactase production into adulthood). But there is such a mechanism (to limited degree) for those who are genetically intolerant.

  123. Aft says:
    @res

    Partly this:

    A sweeping review of shifts in the labor force since 1960 suggests that a sizable portion of the national weight gain can be explained by declining physical activity during the workday. Jobs requiring moderate physical activity, which accounted for 50 percent of the labor market in 1960, have plummeted to just 20 percent.

    The remaining 80 percent of jobs, the researchers report, are sedentary or require only light activity.

    https://well.blogs.nytimes.com/2011/05/25/less-active-at-work-americans-have-packed-on-pounds/

    Partly this:

    Animal and human studies show that food consumption increases when there is more variety in a meal or diet and that greater dietary variety is associated with increased body weight and fat.

    https://www.ncbi.nlm.nih.gov/m/pubmed/11393299/

    Partly this:

    Partly this:
    https://ehp.niehs.nih.gov/doi/full/10.1289/ehp.120-a62

    • Replies: @res
  124. res says:
    @Aft

    Thanks. I can’t find the link right now, but I think a while back we talked here about how efforts to unbias the SAT tended to make it a worse predictor of college grades. Presumably because of your: “Certainly some mix of conscientiousness and interest in written school work.”

    I’m not sure how similar “unbias” was to “control for g” (well, actually the inverse, “extract g”), but I don’t think they were identical.

  125. Factorize says:
    @Aft

    Aft, thank you for the urls. What I think is even more interesting is the modern research that starts to unravel the genetics involved. PMID: 28858769, 21980412, 19461999 etc.. This gives a very nice clean feel that this is pure science. When I watched the video with the babies, I wondered whether this was truly an unbiased result. In some of the videos the walking babies seemed to be carried more and in others they were let to squat. It almost makes me wonder whether even with this type of research there is site to site variability that might invalidate the findings. Yet, with the genetic research, it is not obvious to me how one could “put a thumb on the scale” when typically everyone would be blind to the genotypes involved.

    In my primary school it was pretty much a sociological test of “Europe is one big happy family of white people”. It didn’t work to plan. The kids socially fractionated right down ethnic boundaries and often later moved away from the neighborhood to live in their own ethnic enclave in our city. The school was eventually closed and reopened only for children from a single European ethnic group. Apparently even kindergarteners can figure out that everyone is not the same; the alternative conclusion that they are bigoted would be a tough sell when looking at photos of them essentially being happy kids who like people that are just like them. That is the counter-offensive that I am trying for against sociological tabula rasaism.

  126. CanSpeccy says: • Website
    @res

    Re: Torturing the data

    So you take data for just 13-year-olds — an age at which there will be significant differences in sexual maturation, which are known to impact mathematical ability (see Table 1 here); then take the data for the top 1% and throw away the rest; and finally, group what data you have retained in just four bins.

    You can conceal a lot with that kind of analysis.

    As for the effect of developmental differences, Malcolm Gladwell covers that somewhere, i.e., how kids who develop first get picked for extra training and fast-track development. Even when the late developers catch up in size, they never catch up in competitive success.

    Gladwell was talking mainly about size and strength and its impact on sports training, but differences in developmental age must have similar effects on academic progress and training too.

    • Replies: @Aft
    , @Aft
    , @res
  127. Aft says:
    @CanSpeccy

    Ah, Gladwell, the sad truth is he actually was born with a blank slate.

    http://www.unz.com/isteve/gladwell-at-his-gladwelliest/

    (The effect mentioned matters for sports and somewhat for social dominance; if you want to pretend something “must” be true why not learn to use Google instead of arguing by silly analogies about something that’s almost certainly been researched.)

  128. Aft says:
    @CanSpeccy

    Re: Torturing the data

    So you take data for just 13-year-olds — an age at which there will be significant differences in sexual maturation, which are known to impact mathematical ability (see Table 1 here); then take the data for the top 1% and throw away the rest; and finally, group what data you have retained in just four bins.

    You can conceal a lot with that kind of analysis.

    You actually can’t really conceal anything with that type of analysis.

    Those are quartiles (not cherry-picked cutoffs). It’s monotonic (you can google what that means). The odds ratios are enormous.

    They’re the top 1% because giving the SAT to children any duller than that at 12 years of age is a form of cruel and unusual punishment. (Duke TIP and SMPY have thresholds… No data was “thrown away”)

    And for the more dense, the point of the chart is that the “IQ doesn’t matter beyond 120/130” argument holds exactly no water.

    As to puberty–were you attempting to make an argument or just very obliquely pointing out what’s long been known:

    in selecting any child testing far up in the top one per cent—say at 160 IQ or above (100 IQ being par)—there is far more than an even chance of having thus automatically selected a tall, healthy, fine-looking, honest, and courageous child, with a great love of adventure and of beauty in his makeup

    – Leta Hollingworth (possibly female… 😉

    https://www.davidsongifted.org/Search-Database/entry/R16186

    • Replies: @CanSpeccy
    , @res
  129. res says:
    @Aft

    All good points. I wasn’t familiar with the research on variety. It looks like that gets discussed together with palatablity and (macro) nutrient density.

    Here is a paper which talks about all three of those together: Variety, Palatability, and Obesity
    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4224225/

  130. res says:
    @CanSpeccy

    Developmental differences are worth worrying about. Especially if the sample is racially diverse (which I am guessing is not really the case here). Does anyone know what kind of longitudinal IQ data the SMPY has on its subjects? Presumably they at minimum have SAT (normal age) and GRE type data.

    Regarding “throw away the rest”, on the contrary, you can compare to population level averages.

    The four bins are a pretty standard analysis technique (but see my upcoming response to Aft, there is a subtlety here). When I do analysis like this I like to do quartiles or quintiles (deciles if enough data) but also try to do a spline fit to assess the overall shape of the curve.

    I don’t think you have really supported your “torturing the data” assertion, but leave it to others to decide that for themselves.

    P.S. The point you take from Gladwell is one where I think he is on target (would be interested in hearing a real counterargument with evidence though), but given the overall credibility of his work (e.g. the 10,000 hour rule) would cite him with caution and at least an attempt at backup. Here is an example paper: https://onlinelibrary.wiley.com/doi/abs/10.1111/sms.12002
    There is discussion of this for both sports and academia at https://en.wikipedia.org/wiki/Relative_age_effect

    I was surprised to find the effect was negative for academics (too focused on my own situation, I guess). I was young for my grade and found it a useful challenge which pushed me to excel, but I suppose that near the boundaries it just means someone finds it harder to compete successfully.

    • Replies: @CanSpeccy
    , @res
    , @CanSpeccy
  131. CanSpeccy says: • Website
    @Aft

    were you attempting to make an argument

    (you can google what that means)

    if you want to pretend something “must” be true why not learn to use Google instead of arguing by silly analogies about something that’s almost certainly been researched.

    And for the more dense,

    All of which says, be wary of arguing with IQ-ists. They may, like Aft, have no respect the rules of productive debate, opting instead for the relentless use of insults. That surely says something about the validity, if any, of their arguments, if any (as opposed to sweeping assertions).

    As for density, I do check all the boxes on Res’s chart, plus some not included.

    • Replies: @Aft
  132. CanSpeccy says: • Website
    @res

    Will look at your references with interest. Thanks.

  133. res says:
    @Aft

    Those are quartiles (not cherry-picked cutoffs). It’s monotonic (you can google what that means). The odds ratios are enormous.

    All of those are true, but there is an important (I think) subtlety here which doesn’t get mentioned enough IMHO. Reproducing Figure 1 and its caption here:

    Notice that it groups Cohorts 1, 2, and 3. Also notice how far out the mean SAT-M score for the highest quartile is compared to the others. I don’t think that is what you would expect doing an analysis like this for the top quartile of a randomly picked sample from a normal distribution. And there is a reason for that. I talked about this three years ago in this iSteve comment: http://www.unz.com/isteve/cnn-math-is-racist/#comment-1561520
    Reproducing part of that comment:

    Cohort N; Years when identified; Students’ age when identified; Identification criteria
    1 2,188 1972–1974 12–13 SAT-M 390 or SAT-V 370 (top 1 in 100)
    2 778 . 1976–1979 12 SAT-M 500 or SAT-V 430 (top 1 in 200)
    3 501 . 1980–1983 12 SAT-M 700 or SAT-V 630 (top 1 in 10,000)

    The original of the figure in Nature notes: “Participants from Study of Mathematically Precocious Youth (SMPY) Cohorts 1, 2, and 3 (N = 2,385) are separated into quartiles based on their age-13 SAT-M score.”
    But I did not see any detail of how participants from the first three cohorts were selected for presentation in the figure (notice N from summing the cohort data above is 3467).

    In any case, the “quartiles” are clearly not 1%, 0.75%, 0.5%, 0.25% (as I originally naively expected). There is a large overweighting of 0.01% in Q4 which helps explain the larger difference there.

    So I actually think those odds ratios overstate the effect given how a naive observer would interpret the graphic and its quartiles. To a large (but not total) degree they are comparing the top 0.01% to the 1-0.1% group.

    And “possibly female” ; )

    https://en.wikipedia.org/wiki/Leta_Stetter_Hollingworth

    P.S. Some more discussion of SMPY and Anne Roe at http://www.unz.com/jthompson/your-iq-in-2-minutes/

    P.P.S. To any “anti-IQists” inclined to respond to my quartile point above with howls of outrage I ask two questions:
    1. Why didn’t one of you figure this out and raise this issue?
    2. If I actually conform to the stereotype you keep assigning to me, why did I do it?

    • Replies: @Aft
    , @Aft
  134. res says:
    @res

    I was thinking about the relative age effect some more and thought it was worth expanding that discussion a bit. In particular, how does one reconcile the following two points.

    1. A relative age effect exists for academics and demonstrates that being young for grade negatively impacts long term performance.
    2. Grade skipping has beneficial outcomes for gifted children.

    First, let’s revisit that Wikipedia page for the RAE: https://en.wikipedia.org/wiki/Relative_age_effect

    In academia
    Cut-off dates for academic cohort structuring, including the setting of academic years, are usually determined by national education authorities and tend to be based on autumn start dates, so August or September cut-off dates are common in the Northern Hemisphere and February or March cut-off dates are common in the Southern Hemisphere. This tendency reflects the historical need for children to be involved in summer-time agricultural work with school starting after harvesting.

    A relative age effect in academia is illustrated in the third graph which shows the percent deviation from month of birth profile norms evident in graduations from Oxford University over a 10-year period. Academic relative age effects seem to be moderated by culture.[11]

    A 2006 study finds that relative age affects student performance and has long-lasting effects on life outcomes. The authors find that “the youngest members of each cohort score 4–12 percentiles lower than the oldest members in grade four and 2–9 percentiles lower in grade eight… data from Canada and the United States show that the youngest members of each cohort are even less likely to attend university.”[12]

    A 2014 study finds that Italian students born in the early months of the year “are more likely to be tracked in more academic schools rather than in vocational schools.”[13]

    First, I think they include one piece of “evidence” which is nothing of the sort. The first half of the quote in the third paragraph compares older to younger individuals and is IMO laughable (unless the sole point is to quantify the difference observed). The second half of that quote makes the important point about resultant university attendance though. And the points in the other paragraphs look sound to me.

    Before going further, let’s allow some estimation of how intelligence variation within age compares to variation longitudinally. This page and graph are a great resource for that: http://mindsbasis.blogspot.com/2016/03/rasch-measure-of-intelligence-age-2-25.html

    Eyeballing that graph we see that at 4 years old +1 SD is similar to the average about 1.5 years older. At 8 years of age the curve has flattened out and +1 SD is similar to the average about 5 years older (that age difference is shocking to me).

    Also note that a typical top track might be +1SD and up (1/6 of class) and a typical G&T class might be +2SD and up (top ~2%).

    So thinking about all of this, I see the following possible reconciliations. First, a simple idea: skipping a grade singles someone out positively. Even if they are only average with the older children that is still exceeding expectations.

    Second, given sufficient advantage skipping grades is beneficial. Note from the shape of the curve above the necessary advantage changes over time (greater threshold when younger). The gifted research and typical G&T threshold actually suggest a threshold of +2SD, which sounds plausible.

    A sports analogy might be useful. Consider the case where making a traveling team is a key threshold for advancement and put that at +2SD (~2%) average and +1SD (~16% so 1 in 6 people play and 1 in 8 players make the traveling team) for people who play at all. Further speculate that being a full year younger means a -1SD disadvantage in ability. This means that the youngest needs to be +3SD (~1 in 1000) rather than the +2SD (~2 in 100) needed for the older. Obviously that has an immense impact on chances for advancement. The question is how well that maps to academics. My guess from the curve above is that the academic advantage per year of age is smaller than the corresponding athletic advantage (especially around puberty!). Does anyone have real data about this? Or contrary opinions?

    So to my mind the threshold for grade skipping will depend on other opportunities available. My guess would be you want to have enough advantage to still be able to participate in the most advanced programs for the grade ahead. Does this seem plausible as a rule of thumb?

    An interesting question is whether to use different thresholds at different ages (based on curve above). I would guess not, and just rely on the curve flattening out by age 8. Though that may be different in an area with early G&T (say NYC’s extremely competitive kindergarten G&T admissions).

    Does anyone know how college admissions offices take grade skipping into account when looking at grades and test scores? That seems potentially important.

    Any other ideas for reconciling those initial two assertions?

    P.S. This is a bit jumbled, but hopefully it is detailed and clear enough to prompt some discussion.

    • Replies: @Aft
  135. CanSpeccy says: • Website
    @res

    Yes, “torturing the data” was not the best way of indicating what I meant, which was that there is a universe of actual or potential data out there of which the authors show only narrow slices of what has already be diced and sliced.

    Thus, for example, among grade school kids, they consider only a one twelfth slice by age, an age at which developmental effects are likely most significant.

    As for data binning, that’s fine to illustrate certain points but I’d always like to see the cloud of data, for what it might suggest — something perhaps quite different from what the authors have settled on.

    Then there’s the selection of recorded achievements. How different might the results appear if different criteria were employed? For example, how many of those mathematically precocious kids in the bottom quartile of the 1% finished up with prestigious university chairs in, say, history or biology? And how many of the top quartile ended up as bums, janitors, or fishermen.

    In other words, you could look at the same universe of potential data and get a thousand different impressions.

    One thing, it occurs to me, one should ask about the data is how come those top quartile one percenters are so precocious?

    Is it really just innate superiority? I very much doubt it. G.H. Hardy found math for Cambridge undergraduates boring since he’d covered the curriculum before reaching the university. Thing is though, he’d been a student at Winchester, where children received one on one coaching in subjects that appealed to them.

    A kid I know, likewise received such coaching from a mathematically-gifted undergraduate Brahmin from New Delhi only to find, like Hardy, that undergraduate-level math was uninteresting. Kids with that kind of advantaged background will surely show superior mathematical reasoning ability than a kid whose only exposure to math is in school classroom, which indicates that socio-economic status and home cultural background has a huge impact on the academic performance of 13-year-olds. What’s more, those same factors must go a long way to explain an individual’s career achievements. In other words, the data, leave wide open the question of how much math competence and career achievement are due to innate factors and how much to social factors.

    • Replies: @res
  136. res says:
    @CanSpeccy

    Yes, “torturing the data” was not the best way of indicating what I meant, which was that there is a universe of actual or potential data out there of which the authors show only narrow slices of what has already be diced and sliced.

    That’s fair. But doing just that is important for effective presentation. The big question is whether it is done to clarify and enlighten or mislead and obscure. I think the SMPY researchers tend to do the former. It is just that their priorities are very different from yours.

    In other words, you could look at the same universe of potential data and get a thousand different impressions.

    There is a vast literature concerning the SMPY if you really want to dig deeper:
    https://www.gwern.net/SMPY

    The Anne Roe bibliography there might also be of interest:
    https://www.gwern.net/SMPY#anne-roe

    Regarding (emphasis mine):

    Is it really just innate superiority? I very much doubt it.

    I agree with that. Though I think we would disagree on the precise point between 0% and 100% implied by “not just.”

    One thing which makes it hard to evaluate the effect you are describing is the tendency for individuals to create (at minimum heavily influence) their own environments as they get older.

    The thing is, the issue you are focusing on is incredibly important for maximizing potential. Which to my mind is also one of the best reasons to support IQ research. I think the talent identification aspect of IQ tests was an important advance. By no means are they a complete solution, and they can introduce a different sort of error than what happened before, but net overall I think they are beneficial. This last point is an area where we seem to disagree strongly, but I am not sure exactly why.

    Can you suggest a research study design which would improve on how we look at the nature/nurture aspect compared to how we do so now? And how we might use that knowledge to improve talent development and identification?

    • Replies: @CanSpeccy
  137. utu says:

    Can you suggest a research study design which would improve on how we look at the nature/nurture aspect compared to how we do so now?

    You guys need Dr. Mengele. Say, you can take 100 Jewish newborn and let them be stolen by Gypsies.

    • Replies: @mikemikev
  138. mikemikev says:
    @utu

    That made me wonder how black Africans score in Israel among the hardcore racial egalitarian Jews. Presumably the racist hate of North Europeans somehow beams its way in and depresses them into stupidity.

  139. CanSpeccy says: • Website
    @res

    Speaking of “reasons to support IQ research,” I would say that there are none. IQ-ism is just a phase in the development of psychology as a pathological intellectual discipline. IQ-ism is the latest in a series of attempts to comprehend the vast complexity of the operation of the brain by alchemically simplistic means.

    First was psychoanalysis, aptly described by Peter Medowar as:

    … like a dinosaur or a zeppelin; no better theory can ever be erected on its ruins, which will remain for ever one of the saddest and strangest of all landmarks in the history of twentieth century thought.

    Then there was Behaviorism, which sought to explain human behavior and personality in their entirety solely in terms of the acquisition of operant conditioned reflexes. That theory crashed and burned as cybernetics confirmed what Behaviorists had denied, namely, that humans are conscious beings and that what consciousness tells of our feelings and intentions is a valid source of information.

    And now we have IQism, which claims to be able to quantify a person’s intelligence on a unidimensional scale by means of a simple paper and pencil test involving a few logical puzzles plus, depending on the test of choice, miscellaneous other items.

    How do the IQ-ists sell this idea? Primarily by the artful use of language. Their little test, they call an “intelligence test,” thereby establishing in the minds of the masses the unquestioned assumption that intelligence is what the IQ-ists test measures. In fact, however, as a Google search will confirm, intelligence is the ability to acquire and to use information, whereas an IQ test measures neither except in an incredibly limited domain and with a test the results of which are subject to massive circumstantial bias.

    But the IQ-ist scam has worked so well for so long that psychology has yet to even broach the real scientific questions that must underlie the measurement of intelligence: namely, how to measure the capacity for information acquisition; and how to measure skill, effectiveness, Darwinian fitness, or whatever, in the use of information.

    When one considers the measurement of intelligence in those terms, one is confronted by the realization that information is acquired via multiple channels, auditory, olfactory, visual, proprioceptive, etc. with data from each channel processed by a specialized brain modules, or probably in most if not all cases, by multiple specialized brain modules.

    So now if we take account of the fact that there are hundreds if not thousands of structural genes that impact the development and characteristics of those sensory channels and processing modules, we see that the capacity for the acquisition of information is not a single characteristic of the brain but a large collection of independent variables, as is well known to common sense. People vary hugely in powers of memory and, moreover, that variation is type specific. Mozart transcribed the entire Allegri miserere after a single hearing, Stephen Wiltshire sketched the whole of Red Square from memory after a brief visit. Mozart, so far as we know, was no graphic artist, and Stephen Wiltshire is no musician. Others do more or less brilliantly remembering poetry, the numbers of pi, or conversational tittle tattle.

    So in only the matter of data acquisition, we see that intelligence is multiple not unitary. But much more complex to analyse than the capacity for information acquisition is the capacity for the use if information. In fact, perhaps, that is an impossibly difficult challenge. But it is a challenge that must be faced by anyone who claims to measure intelligence in a scientific and quantitative way.

    As for the innateness of intelligence, it is axiomatic that the potentiality is entirely innate. Moreover, we know that there are hundreds, perhaps thousands of genes that direct brain development, plus probably many thousands of hereditary controlling elements, most yet to be identified, that shape the development of the brain and hence intelligence.

    But the function of the brain is to record both sensory inputs, i.e., experience, and the internal workings of the brain, i.e., the development of our ideas, both of which shape the way we use information. So certainly, environmental factors that shape the contents of mind must have a huge impact on intelligence, which is why focusing the genetic basis of intelligence to the exclusion of environmental factors, such as education and culture, cannot result in a comprehensive understanding of the phenomenon.

    • Replies: @mikemikev
  140. mikemikev says:
    @CanSpeccy

    Nice start poisoning the well with some irrelevant comparisons. For other anti-science attacks they have to resort to phlogiston and flat earth, but you have those. We can happily agree that (((psychoanalysis))) is garbage.

    You’re saying that intelligence is caused by multiple components, therefore intelligence itself is multiple. It’s like saying a car’s top speed is multiple because there are spark plugs and pistons. It doesn’t seem to make any sense.

    • Replies: @CanSpeccy
    , @CanSpeccy
  141. CanSpeccy says: • Website
    @mikemikev

    You’re saying that intelligence is caused by multiple components, therefore intelligence itself is multiple. It’s like saying a car’s top speed is multiple because there are spark plugs and pistons.

    Actually I didn’t say that. Perhaps intelligence tests should include an element for reading comprehension.

    • Replies: @mikemikev
  142. mikemikev says:
    @CanSpeccy

    Actually I didn’t say that.

    Oh yes you did.

    • Replies: @CanSpeccy
  143. CanSpeccy says: • Website
    @mikemikev

    Oh yes you did.

    It does seem that some of the firmest supporters of IQ-ism have comprehension issues.

    What I said was that if you want a quantitative measure intelligence, which is the capacity to acquire and use information, then you need to quantify the ability both to acquire information and to use information (in ways such as I specified). I further said that information acquisition is via multiple channels and involves processing by multiple brain modules. Therefore, that part of intelligence that consists in information acquisition depends on multiple components of the central nervous system, meaning that there are multiple independently variable (due to either genetics or environmental factors) components of the central nervous system upon which that aspect of intelligence (information acquisition) depends.

    In turn, what that means is that intelligence is not unitary, but has many independently varying components. And that means that intelligence cannot be measured by a single number, whether called an IQ or anything else.

    • Disagree: mikemikev
    • Replies: @CanSpeccy
    , @mikemikev
  144. CanSpeccy says: • Website
    @mikemikev

    Nice start poisoning the well with some irrelevant comparisons.

    There was nothing irrelevant in my mentioning the now discredited psychological schools of Freudianism and of Behaviorism. The prior emergence of those ideas provides evidence in support of my contention that psychology is prone to pathological deviation from the scientific method: IQ-ism being the third manifestation of this tendency within a century.

  145. CanSpeccy says: • Website
    @CanSpeccy

    Disagree: mikemikev

    You mean you disagree that that is what I said? I disagree with that!

  146. Logan says:
    @RaceRealist88

    This sounds remarkably like what got Galileo in trouble. He had no underlying theory to explain why the earth went around the sun, therefore his observations that it did must be in error.

  147. mikemikev says:
    @CanSpeccy

    What I said was that if you want a quantitative measure intelligence, which is the capacity to acquire and use information, then you need to quantify the ability both to acquire information and to use information (in ways such as I specified).

    Yes that’s what IQ does.

    I further said that information acquisition is via multiple channels and involves processing by multiple brain modules. Therefore, that part of intelligence that consists in information acquisition depends on multiple components of the central nervous system, meaning that there are multiple independently variable (due to either genetics or environmental factors) components of the central nervous system upon which that aspect of intelligence (information acquisition) depends.

    In turn, what that means is that intelligence is not unitary, but has many independently varying components.

    It’s doesn’t matter how many components cause intelligence for intelligence to be largely captured by a single factor.

    • Replies: @CanSpeccy
  148. Aft says:
    @CanSpeccy

    700+ SAT Math at age 12, have a STEM doctorate and STEM Tenure at a top 50 university, and some patents too? 😏

    • Replies: @CanSpeccy
  149. Aft says:
    @res

    I remember looking into that once (because the 700 is way too high).

    Other source: https://www.gwern.net/docs/iq/smpy/2018-lubinski.pdf

    Either way it refutes all the “high IQ, your income goes down, you too smart, you do IMO” trolls. Income is still increasing. Monotonic all the way. And other accomplishments in STEM are increasing even more materially.

    • Replies: @res
  150. Aft says:
    @res

    Good catch.

    I remember pondering on that once. He has another paper that uses total scores and still he seems unclear on what quartiles mean…https://www.gwern.net/docs/iq/smpy/2009-lubinski.pdf

    Main point is proven either way: things are monotonic the whole way. Income goes up. Stem type accomplishments go up even more.

    • Replies: @res
  151. Aft says:
    @res

    Very different being an average kid than way smarter than everyone around you. Your intuition sounds right: it’s a plus to avoid boredom and actually stay engaged if very bright, but not so great for the average or slow kids who are also a year young taking the SATs, starting college, everything.

    Very interesting curves from 18-25.
    And before that too…

    Not a lot of people skip grades these days and it’s probably a bad idea unless well beyond 130.

    For beyond that, the best idea I’ve seen was Leta’s final conclusion: start high school at 13 to be socially matched. But middle school is a total waste so best to study orthogonal subjects (languages, the world, society, biographies, or anything you want) on your own during those years and then rejoin at high school time for the “college preparatory curriculum” that’s still trivially easy but at least cognitively sorted to be around IQ 120+ types mostly.

    • Replies: @res
  152. res says:
    @Aft

    Thanks! Since there are so many SMPY papers and so little time I appreciate specific recommendations.

    Figure 15.1 is interesting in that it is similar to, but subtly different from Figure 1 above. Figure 1 states “Adapted in part from Park, Lubinski, and Benbow (2007, 2008)” while Figure 15.1 states “Adapted from Lubinski (2009).”

    Figure 15.1 is based on SAT M+V composite (Figure 1 is SAT-M only) with the quartile means being 858, 966, 1056, and 1231. The cut scores give good perspective on the quartile means in Figure 1 (from the caption): “cutting scores for the top 1 in 200 were SAT-M >= 500, SAT-V >= 430; for the top 1 in 10,000, cutting scores were SAT-M >= 700, SAT-V >= 630 by age 13.” So referring back to Figure 1 we see that the Q2 mean is over the 1 in 200 cutoff and the Q4 mean is right around the 1 in 10,000 cutoff.

    Figure 15.2 gives an interesting look at typical ability level and tilt for some of the accomplishments. One intriguing wrinkle was the difference between top-50 and below tenure track academics. For the humanities the top 50 group had slightly more ability and a smaller (more towards Q!) tilt while for STEM the tilt was slightly less towards Q but with higher ability. Another interesting comparison is JDs and MDs had similar ability to <PhD terminal humanities and STEM, while humanities/STEM PhDs were significantly higher and close to both faculty and publication levels.

    Figure 15.3 has scatterplots for both the SMPY and Duke TIP (0.01% cutoff groups) along with an estimated IQ 160 line. I wish they had included some other IQ lines to give an idea of the variation within the group.

    Figure 15.4 gives mean levels for various creative accomplishments. One nice feature of these graphs is the axes give both Z-scores and raw SAT subtest scores. If I interpret the plot correctly they also give the number of subjects having each achievement.

    The latter part of the paper talks about spatial ability. This caught my eye.

    A discriminant function analysis employed the age 13 mathematical, spatial, and verbal ability assessments to predict the four types of creative outcomes described previously 35 years later. Mathematical and verbal ability scores jointly accounted for 10.5% of the variance in creative group outcomes. The inclusion of spatial ability — the “orphan ability” — added another 7.5% to that variance.

    Also this:

    Unfortunately, approximately half of young adolescents in the top 1% in spatial ability are missed by modern talent searches restricted exclusively to mathematical and verbal reasoning ability (Wai et al., 2009; Wai & Worrell, 2016). This omission not only neglects an underserved population — and a critical source of human capital for technical professions — it also constitutes a lost opportunity for the kind of refinements seen in Figures 15.5 and 15.6 for all individuals. Individuals may be highly similar on any twospecific abilities (mathematical/spatial/verbal), but if they differ markedly on the third, differential development can be anticipated.

    Perhaps this excerpt will cause CanSpeccy to reconsider how accurate his IQist strawman is? (probably not, but worth including anyway IMHO)

    The challenge for educators and career counselors is to find the optimal niche for each student, so that he or she can maximize the positive aspects of his or her individuality (Lubinski, 1996, 2016; Lubinski & Benbow, 2000). This is best done by knowing and treating each student as an individual.

    This referenced paper looks like a good read for anyone interested in gifted education: https://www.gwern.net/docs/iq/smpy/1996-benbow.pdf
    I like the title: Inequity In Equity: How ‘Equity’ Can Lead to Inequity for High-Potential Students

    Another quote for CanSpeccy:

    More than 60 years ago, Lewis Terman (1954) reflected on his multiple decade longitudinal study by affirming the importance of initially using general intelligence to identify participants when the gifted field was young and his groundbreaking work had just begun. Nevertheless, he then added: “[s]uch tests do not, however, enable us to predict what direction the achievement will take, … both interest patterns and special aptitudes play important roles in the making ofa gifted scientist, mathematician, mechanic, artist, poet, or musical composer, …” (p. 224).

    Agreed that this is a key point of the SMPY data:

    Either way it refutes all the “high IQ, your income goes down, you too smart, you do IMO” trolls. Income is still increasing. Monotonic all the way. And other accomplishments in STEM are increasing even more materially.

    P.S. The OCR in that paper was a bit rough. Sorry for any errors I missed (or made).

    • Replies: @James Thompson
  153. res says:
    @Aft

    That reference was also interesting. Thanks. The title serves as a good indicator of the focus: Cognitive epidemiology: With emphasis on untangling cognitive ability and socioeconomic status
    Won’t go through it in as much detail, but this was an interesting point.

    The GCA/SES confound may be reasonably addressed in the following way. Using a large stratified random sample of the U.S. 10th grade student population secured by Project Talent (Flanagan et al., 1962), N=95,650 participants, Lubinski and Humphreys (1992) selected the top 1% on a measure of cognitive ability, for each sex, as well as the top 1% on a measure of SES. The four resulting groups, gifted boys n=497, gifted girls n=508, environmentally privileged boys n=647, environmentally privileged girls n=485, had minimum overlap. Only 41 boys and 46 girls were members of both the privileged and gifted groups.

    I find that low overlap surprising given the IQ/SES correlation. Any thoughts? The correlation in adolescence seen in https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3270016/ was 0.37. I should try a simulation to see how well that matches the observed <10% overlap. Also, I wonder what is going on that there were a third more environmentally privileged boys than girls?

    • Replies: @Aft
  154. res says:
    @Aft

    Very different being an average kid than way smarter than everyone around you. Your intuition sounds right: it’s a plus to avoid boredom and actually stay engaged if very bright, but not so great for the average or slow kids who are also a year young taking the SATs, starting college, everything.

    I think there are some useful conclusions to be drawn about when to start school and/or skip/hold back grades for students of differing abilities. Of course, physical and emotional maturity are also important. As an example, accelerating someone too much might make it hard to compete in sports.

    Very interesting curves from 18-25.
    And before that too…

    There is some good discussion of those curves and some related ideas (and references) from commenter EH in Dr. Thompson’s archives. This fairly recent thread provides one example, but there are more if you look: http://www.unz.com/jthompson/origins-of-iq-tests

    • Replies: @Aft
  155. CanSpeccy says: • Website
    @Aft

    Dunno about SAT Math at age 12, although I was, I seem to recall, top of my class. As for the rest, yes, although I resigned the tenure-track university appointment after three days. Phew! What relief! That University currently ranks in the top 30, worldwide (Times Higher Ed). The patented technologies weren’t worth much, although one has been in continuous use internationally for over 40 years. However, unpatented technology that I developed saved taxpayers tens of millions and quite likely hundreds of millions.

    But why am I talking about this? I thought we were talking of the reality of g. And yes g is real, mine goes up and down depending on whether I had breakfast this morning, flew the Atlantic last night, or just downed a glass of champagne. But that absolutely does not mean that because I fail to understand Byzantine Reliable Broadcast or the intricacies of quantum teleportation, that there are not some things I understand better than most people (and other things I likely understand less well than most people).

    And that is my point. Intelligence is not unitary.

    Let me say that again. Intelligence is not one thing. It cannot be measured by a single number on a unidimensional scale.

    The concept of IQ is not only bollocks, but fiendishly evil in its consequences. It discourages those who score poorly from trying harder. It discourages those who score well to try with all their might. It discourages recognition of accomplishment as opposed to supposed innate gifts. And it undermines the incentive to discover the genius of every individual, however, overall dumb we may seem to be.

    • Replies: @mikemikev
  156. CanSpeccy says: • Website
    @mikemikev

    It’s doesn’t matter how many components cause intelligence for intelligence to be largely captured by a single factor.

    To quote another commenter: Sigh…

  157. mikemikev says:
    @CanSpeccy

    Let me say that again. Intelligence is not one thing. It cannot be measured by a single number on a unidimensional scale.

    That’s wrong.

    • Replies: @Anonymous
  158. @res

    My take was, as yours, that the spatial element had been left out of the early talent search work, and might be partly responsible for missing some very productive thinkers.

    • Replies: @res
    , @dux.ie
  159. res says:
    @James Thompson

    I was surprised how much variance explained (7.5%) spatial added to math and verbal together (10.5%). It would be interesting to see how well each did as a single variable predictor since they are all correlated. Also interesting to compare those to a derived g (both with and without a spatial component).

    I also wonder how much variance is explained by other variables they have (e.g. SES) either alone or in conjunction with the test variables.

    P.S. Was SMPY the first talent search to look at spatial ability? How much was it considered before Anne Roe’s work?

    • Replies: @James Thompson
  160. Anonymous[306] • Disclaimer says:
    @mikemikev

    CanSpeccy is a resident anti-IQ moth who can’t stop himself from hitting a nearby IQ bulb.

    Most commenters learn to ignore him on this topic.

    • Agree: Aft
    • Replies: @mikemikev
  161. CanSpeccy says: • Website

    Here’s a figure from a study on the effect of volatile organic compounds and carbon dioxide on cognitive performance at a range of concentrations that commonly occur in air conditioned office buildings:

    What is shows is that the effect of to exposure to organic air pollutants and carbon dioxide on cognitive performance varies greatly according to the cognitive task measured. This confirms, (a) that cognitive capacity are not constant but subject to large variation according to immediate environmental factors, and (b) that intelligence is not unitary but involves multiple capacities that vary independently under the influence of environmental stress.

    In particular, note the catastrophic effect of carbon dioxide at concentrations that are common in school and office buildings on strategic thinking compared with its minimal effect on some other tasks such as information seeking, etc.

    • Replies: @mikemikev
    , @res
    , @Aft
  162. @res

    Jensen pointed out, I forget where, that g was stronger at lower levels of ability. It is still there at the highest levels, but testing over a wider range of domains is worth while if you want to make fine-grain predictions.

    • Replies: @res
  163. mikemikev says:
    @Anonymous

    I just wondered how he’d react to someone mirroring his ad nauseam flat-contradiction assertions.

  164. mikemikev says:
    @CanSpeccy

    You’re saying people suffer cognitive impairment if you put them in a room filled with carbon dioxide? I’m not sure anyone has taken this into account.

    • Replies: @CanSpeccy
  165. res says:
    @James Thompson

    Thanks! I followed up on this and Jensen discusses it in The g Factor (any better references welcomed). From page 226:

    There is some independent evidence that at lower levels of IQ, more of the test variance is attributable to g than to group factors, as compared with higher levels of IQ. This phenomenon, which Spearman dubbed the “ Law of Diminishing Returns,” is more fully discussed in Appendix A. It would be of considerable theoretical importance if it were firmly established as generalizable to all test batteries.

    From page 321:

    Spearman’s “ Law of Diminishing Returns” (see Appendix A, pp. 585-88) states that less of the variance in a collection of diverse mental tests consists of g within a high-ability group than it does within a low-ability group. A corollary of this “ law” is that the average intercorrclation among diverse tests is smaller for a high-ability group than it is for a low-ability group. The mounting evidence that this is an authentic phenomenon suggests a possible test of the hypothesis that the secular increase in IQ involves to some extent an increase in the actual level of g in the population. It would be supported by finding a secular decrease in the average intercorrelation among tests (and hence a decrease in their g loadings).

    Some excerpts from Appendix A.

    Spearman (1927, pp. 217-21) compared the disattenuated correlation matrices (based on 12 diverse cognitive tests) of 78 “ normal” children and 22 “ defective” children. He found that the mean r of the matrix for the normal children was +.466; for the retarded children the mean r was +.782. Deary and Pagliari (1991) performed principal components analyses of Spearman’s correlation matrix for the normal children and the correlation matrix for the defective children. The average loadings on the first principal component (PCI) of each matrix were +.725 and +.899, respectively. Yet the PCI was clearly the same factor in both the normal and retarded groups, as indicated by a congruence coefficient of +.988. Spearman also noted in other data sets that tests’ intercorrelations (and average g loadings) were larger for younger children than for older children. These findings suggested that the higher the level of g, the less is the amount of g variance in any particular mental test.

    The first really systematic and methodologically convincing study of this phenomenon was conducted by Detterman and Daniel (1989). They demonstrated the effect both with a variety of computer-administered cognitive tasks and with the subtests of Wechsler Adult Intelligence Scale-Revised (WAIS-R) and the Wechsler Intelligence Scale for Children-Revised (WISC-R). The latter demonstration, based on very large subject samples on the Wechsler tests, was particularly impressive. The entire subject sample (on the WISC-R and on the WAIS-R) was divided into five levels of IQ (122), with the IQ equivalent based on only one subtest. In the first analysis the Vocabulary subtest was the basis of classification; in the second analysis, Information. Within each of the five ability levels, the average intercorrelation among all the Wechsler subtests (except Vocabulary or Information) was obtained. These average intercorrelations among subtests decreased monotonically from about +.7 for the IQ < 78 group to about +.35 for the IQ > 122 group.

    Multiple possible explanations are discussed. This one seems plausible to me:

    The result was interpreted in terms of Detterman’s (1987) systems theory of mental retardation. This theory posits that mental ability involves a number of distinct systems or processes, and that some processes are more “ central” than others, in the sense that their functioning is crucial to a wider range of cognitive operations. A deficiency in a highly central process will therefore handicap a great many mental functions and result in low scores on every type of test. Variation in less central functions will affect only certain narrow abilities (group factors and specificity) but not most abilities. Persons with low IQs have less efficient central processes, hence overall low performance on most kinds of cognitive tasks. Persons with higher IQs have more efficient central processes but may vary considerably in the less central, narrower processes. Consequently, there should be higher correlations (and more g variance) among various tests in a low-IQ group and lower correlations (less g variance) in a high-IQ group. A corollary of this theory is that the “ profile” or pattern of subtest scores should be flatter (i.e., lower standard deviation among the individual’s subtest scores) for low-IQ persons than for high-IQ persons.

    Jensen has an extended discussion of how restriction of range impacts detecting this effect.

    The appendix concludes with some interesting but non-conclusive results concerning “Heritability as a Function of IQ Level. “

  166. CanSpeccy says: • Website
    @mikemikev

    You’re saying people suffer cognitive impairment if you put them in a room filled with carbon dioxide?

    You pretend not to understand — or perhaps really don’t understand — the difference between zero point two percent, a common carbon dioxide concentration in school and office buildings, and “a roomful”.

    But that kind of pathetic pseudo-debating point — argument by the childish put-down, is what the world, the US of A anyhow, theoretically the world’s most highly educated country, has largely come to, isn’t it. Losers, incapable of a logical argument upholding some theory about which they understand essentially nothing, by means of the sheer volume of asinine comments.

    But sad to see that kind of contribution on the Unz Review. Also sad to see that, so far at least, the upholders of IQ orthodoxy have had nothing to add to the moron commentary in response to evidence seeming to challenge fundamental assumptions of their ideology: namely, (1) that an individual’s intelligence is a fixed thing, not a variable dependent on environmental factors, least of all transient factors such as air quality; and (2) that cognitive capacities vary relative to one another according to environmental factors.

    But surely someone following this blog about intelligence, will offer an intelligent comment.

    • LOL: mikemikev
    • Replies: @res
  167. res says:
    @CanSpeccy

    Here is the source of that graphic: https://ehp.niehs.nih.gov/doi/10.1289/ehp.1510037

    FWIW, your graphic does not show the higher CO2 cases. It has three cases. The study included 5: Conventional, Green, Green+, Moderate CO2, High CO2. The Green and Conventional cases had half the ventilation rate of the other 3 (achieved by adjusting the ratio of indoor and outdoor air from 50/50 to 0/100). Table 2 summarizes the conditions for each case.

    First, for anyone who cares about sample size, N = 22.

    The TVOCs were >10x (!) higher for the Conventional case compared to all of the others. I have no idea how realistic that is. They go into great detail about which VOCs are present in Table 3.

    One wrinkle is they used their study population as the normative group for their cognitive tests. Not sure how acceptable that is, but it does say something about how well understood their tests are IMO.

    They do seem to use good standards for statistical significance: “Given the multiple comparisons tested in this analysis, p-values < 0.001 were considered to be statistically significant after performing a Bonferroni correction."

    Figure 2 is the graphic showing performance vs. CO2.

    One thing that strikes me as odd is despite all the things they did measure they did not include a direct measure of O2. That seems rather important. Here is a paper looking at supplemental oxygen and cognitive performance in the elderly (see the references for more): https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4107523/

    All in all, an interesting study, but I'm not sure how to interpret the effect sizes given the lack of normative data for the tests.

    It would be interesting to add a comparison case for taking the tests in an open air courtyard surrounded by plants.

    • Replies: @Aft
  168. res says:
    @CanSpeccy

    But surely someone following this blog about intelligence, will offer an intelligent comment.

    Hopefully the comment I wrote before seeing this one of yours will suffice.

    (1) that an individual’s intelligence is a fixed thing, not a variable dependent on environmental factors, least of all transient factors such as air quality; and (2) that cognitive capacities vary relative to one another according to environmental factors.

    Do you really consider these observations so revolutionary?

    The strawmen you assign to others here are tiresome.

    • Replies: @CanSpeccy
    , @CanSpeccy
  169. CanSpeccy says: • Website
    @res

    The strawmen you assign to others here are tiresome.

    You think I am attaching “strawmen” to Mikemikev’s and Anon’s fatuous comments? Perhaps you will elaborate. What I said was that the following were axioms of the IQ-ist thesis:

    (1) that an individual’s intelligence is a fixed thing, not a variable dependent on environmental factors, least of all transient factors such as air quality; and (2) that cognitive capacities vary independently, which negates the fundamental assumption of the IQ-ist thesis.

    But in response to those points, you ask:

    Do you really consider these observations so revolutionary?

    So which are they? Strawmen or facts readily acknowledged by the IQ-ists.

    From what I have read here, these propositions are contrary to the general position of the IQ-ists. But prove me wrong. Give us the examples of IQ-ists presenting evidence that (a) IQ is importantly influenced by environment, and (b) that cognitive capacities vary independently, thus negating the contention that intelligence is a unitary factor measurable by a single number on a uni-dimensional scale.

    • Replies: @res
  170. CanSpeccy says: • Website
    @res

    Re the Allen et al. paper to which I referred (the embedded link that I included in my text seems to have been removed, or possibly I botched it, but thanks for providing an explicit link) you say:

    One thing that strikes me as odd is despite all the things they did measure they did not include a direct measure of O2.

    The reason it would have been pointless to do so is evident if you consider the overall equation for oxidative respiration and the magnitude of the elevation of carbon dioxide concentration in a poorly ventilated office (rarely, if ever, more than 0.2%). The relationship between respired carbon dioxide and consumed oxygen has a 1:1 stoichiometry. Thus, the reduction in ambient oxygen in a stuffy office would, at most, be 0.2%, or one part in 100 versus a five-fold rise in carbon dioxide concentration. So the change in ambient oxygen concentration is physiologically irrelevant.

    But in any case, in the Allen et al. study, carbon dioxide concentration was raised not as the result of respiration, but by direct carbon dioxide injection, so there would have been no appreciable difference in oxygen concentration between the plus or minus elevated carbon dioxide treatment conditions. (These people knew what they were doing, as one would expect of Harvard University researchers).

    • Replies: @res
  171. res says:
    @CanSpeccy

    So which are they? Strawmen or facts readily acknowledged by the IQ-ists.

    The strawman is that they are acknowledged and you pretend they are not. Or play Motte and Bailey with what is acknowledged and how you describe it.

    (1) that an individual’s intelligence is a fixed thing, not a variable dependent on environmental factors, least of all transient factors such as air quality

    Perhaps best expressed as a roughly fixed potential which can be degraded by subpar environment (both short and longer term effects). One interesting question is how much that potential is reduced by the typical Western environment.

    (2) that cognitive capacities vary independently, which negates the fundamental assumption of the IQ-ist thesis.

    The second part of that is your classic strawman. Anyone with any sense acknowledges the first part. The question is the proportions of independent and common variation.

    Give us the examples of IQ-ists presenting evidence that (a) IQ is importantly influenced by environment

    I suspect this is where we get to watch you play Motte and Bailey (alternately, no true IQ-ist).

    I assume you consider Arthur Jensen an IQ-ist? Here is an excerpt from the end of The g Factor Appendix A I mentioned above:

    I have found (Jensen, 1997a), in some fine-grained twin analyses, that favorable and unfavorable nongenetic influences on mental development are not symmetrically distributed, and that lower IQs have a larger component of nongenetic variance due to largely biological environmental prenatal and early childhood influences (e.g., poor nutrition, disease, head trauma, mother-fetus incompatibility in blood antigens, prematurity, and low birth weight). This finding would seem to favor the hypothesis that h2 decreases at lower levels of IQ

    What is it with commenters (and the occasional author, like Chanda Chisala) on the Unz Review handing out such softballs as challenges? Is it just that so few people are actually capable of backing up what they say that those commenters (etc.) get sloppy and start assuming everyone is like that?

    Back to you:

    (b) that cognitive capacities vary independently, thus negating the contention that intelligence is a unitary factor measurable by a single number on a uni-dimensional scale.

    Again, everybody with any sense admits the first part (it’s a matter of degree of common vs. independent variation). And intelligence is measurable by a single number on a uni-dimensional scale. It’s just not a complete measurement.

    It is important to note that there is a significant difference between “environmental influences on IQ exist” and “environmental differences explain a significant portion of IQ variation observed in the population.”

    I would postulate the following environmental influences would be accepted by any intelligence researcher as reducing measured IQ in the moment.
    – Being drunk.
    – Nitrogen narcosis.
    – Excessive noise.
    – Temperature extremes.
    – Sufficiently low oxygen levels.

    P.S. I am a little surprised you did not jump on the g explains less variance for high IQs than low IQs point above. That seems like one of the better arguments against using it for high end selection. Mind you, I think the SMPY makes clear that is not true in an absolute sense, just relative.

    • Replies: @CanSpeccy
  172. res says:
    @CanSpeccy

    You are of course right about the stoichiometry (close enough that I’ll give it to you anyway, in reality it varies in a range from ~0.7-1.0 depending on whether fats/protein/carbs are being burned https://en.wikipedia.org/wiki/Respiratory_quotient ) and the relative atmospheric concentrations of O2 and CO2. And that is probably the reason they chose not to measure O2. I still think it would be useful to record for two reasons:
    – Make certain nothing odd was going on.
    – Quantify the research conditions to make it easier to compare with different altitudes.

    If it was really so hard to measure O2 (while measuring so many other things) barometric pressure would probably be a good enough proxy.

    (These people knew what they were doing, as one would expect of Harvard University researchers).

    As I have mentioned here before, I am fascinated by how you seamlessly switch between extreme skepticism and taking research at face value. Too much of that and it starts sounding like isolated demands for rigor:
    https://slatestarcodex.com/2014/08/14/beware-isolated-demands-for-rigor/

    And if you made that assertion with a straight face you have never looked at Harvard nutritional research from the 1970s. Do a quick search for Frederick Stare’s sugar research. Though thinking about it, I suspect he knew exactly what he was doing. It is just that he had a major conflict of interest.

    “Knew what they were doing” != “Can trust what they wrote”
    As anyone involved with controversial topics (like IQ research, especially on group differences) is (well, should be) well aware.
    (and note that there are different kinds of trust, compare “not lie” with “abstract is an accurate summary of the paper” with “spinning of data in the text”)

    That Allen et al. paper looked good to me (aside from the sample size, which makes me want to see confirmatory research), but do you have any doubt that the conclusion was what they wanted to find? Given that, I think a degree of skepticism is worthwhile.

    • Replies: @CanSpeccy
  173. CanSpeccy says: • Website
    @res

    (2) that cognitive capacities vary independently, which negates the fundamental assumption of the IQ-ist thesis.

    The second part of that is your classic strawman.

    Rubbish. That cognitive capacities vary independently, negates the fundamental assumption of the IQ-ists, namely, that intelligence is unitary and can, therefore, be quantified by a single number.

    What is it with commenters (and the occasional author, like Chanda Chisala) on the Unz Review handing out such softballs as challenges?

    What’s with you that you cannot answer the question except with some out-of-context verbiage signifying nothing.

    Is it just that so few people are actually capable of backing up what they say that those commenters (etc.) get sloppy and start assuming everyone is like that?

    More BS, and no facts.

    I would postulate the following environmental influences would be accepted by any intelligence researcher as reducing measured IQ in the moment.

    Fine diversion by irrelevance. The point about the carbon dioxide effect is that it varies among cognitive capacities, indicating that intelligence is not one thing that can be measured by a single number.

    • Replies: @res
  174. CanSpeccy says: • Website
    @res

    If it was really so hard to measure O2 (while measuring so many other things) barometric pressure would probably be a good enough proxy.

    You think they may have been conducting their experiment as the eye of a hurricane passed over the lab, or what?

    That Allen et al. paper looked good to me (aside from the sample size, which makes me want to see confirmatory research)

    There is confirmatory research:

    This from the Lawrence Berkeley National Lab and State University of New York:

    Is CO2 an Indoor Pollutant? Direct Effects of Low-to-Moderate CO2
    Concentrations on Human Decision-Making Performance

    https://ehp.niehs.nih.gov/doi/pdf/10.1289/ehp.1104789

    As in the later study cognitive capacities were differentially affected, i.e., proved to be independent variables.

    • Replies: @res
  175. res says:
    @CanSpeccy

    Rubbish. That cognitive capacities vary independently, negates the fundamental assumption of the IQ-ists, namely, that intelligence is unitary and can, therefore, be quantified by a single number.

    Now that was a persuasive argument. Do you really not understand that common and individual variation can exist together?

    What’s with you that you cannot answer the question except with some out-of-context verbiage signifying nothing.

    Physician, heal thyself.

    More BS, and no facts.

    Did you miss my Jensen quote? Which was exactly what you asked for.

    Fine diversion by irrelevance. The point about the carbon dioxide effect is that it varies among cognitive capacities, indicating that intelligence is not one thing that can be measured by a single number.

    You have an interesting definition of irrelevance. And an even more interesting standard of proof.

  176. res says:
    @CanSpeccy

    You think they may have been conducting their experiment as the eye of a hurricane passed over the lab, or what?

    I mentioned altitude right before that. Stop being deliberately obtuse.

    There is confirmatory research:

    This from the Lawrence Berkeley National Lab and State University of New York:

    Thanks. What is up with the N=22 in these papers? Is that the number power calculations say is necessary?

    As in the later study cognitive capacities were differentially affected, i.e., proved to be independent variables.

    Sigh. differentially affected != independent
    At least you are right out there with your poor reasoning.

    And before you lash out in response, the definition of independent variables is such that: “If X and Y are independent, then they are also uncorrelated.”
    https://www.stat.cmu.edu/~cshalizi/uADA/13/reminders/uncorrelated-vs-independent.pdf
    (but note that uncorrelated does not prove independence, it is a necessary but not a sufficient condition)

    That the variables move in the same general direction (worse with increasing CO2) indicates they are correlated.

    What’s up here? You are smarter than this. Have you simply run out of good arguments?

  177. CanSpeccy says: • Website

    The trouble with you, Res, is that you have joined a corrupt profession. Psychologists, seeking to uncover the springs of human action and to plumb the human soul, are motivated not solely by the desire for knowledge, but also by the will to power.

    Hence the irresistible attractiveness of IQ-ism: the marketable scam of giving to every individual an intellectual ranking, a single point on a unidimensional scale.

    Hence the brainwashing of entrants to the profession. Here, for example, is the liar, Professor Jordan Peterson, failed LSAT, but claiming an IQ “in excess of 150,” bullying University of Toronto psychology undergraduates:

    One of the things I have to tell you about IQ research is this: if you don’t buy IQ research, you might as well throw away all the rest of psychology. And the reason for that is that the psychologists who developed “intelligence testing” were among the early psychologists that instantiated the statistical techniques that all psychologists use to verify and test all of their hypotheses. So you end up throwing the baby out with the bath water.

    And the IQ people have defined intelligence in a more stringent and accurate way that we have been able to define almost any other psychological construct. So if you toss out the one that is most well defined, then you’re kind of stuck with the problem [of] what are you going to do with all the other ones that you have left over…. whose predictive validity is much less.

    In other words, swallow the IQ-ist bullshit or get out of my class and find another program to pursue.

    What’s up here? You are smarter than this. Have you simply run out of good arguments?

    At least you acknowledge I had good arguments. More than I can say for you. I had hopes of you at one time. But it seems clear now that no reason for hope remains. Though your sophistry is more subtle, like mikemikev and Anon, you have drunk the Coolaid.

    • LOL: res
    • Replies: @res
  178. res says:
    @CanSpeccy

    That was nice. Next time maybe try to respond to my points. In particular regarding your stupidity concerning independent variables.

    At least you acknowledge I had good arguments. More than I can say for you.

    You know, if you can’t credit that I at least occasionally make good arguments here (and again, perhaps you could comment on your misuse of the term “independent variables”? or my correction of your slightly off respiratory stoichiometry comment?) then I think that makes clear exactly what your opinion is worth. You might also take a moment to remember who defended you when some here were calling you stupid (or something like that, I can’t be bothered to chase it down right now). I call things like that as I see them and try to acknowledge the personal strengths and good points made by those I am debating.

    It is really sad when a smart person like you lets his ideology blind him to contrary facts.

    P.S. And a hearty LOL for your failure to admit that I met your challenge in comment 172 (with the Jensen quote in my response). For anyone looking at comment 172 make sure to realize that CanSpeccy screwed up his blockquotes so it can be hard to interpret who said what when.

  179. @The Real and Original David

    Would “regatta” be one of those items?

    How can you expect West Africans to know about Italian pasta sauces?

  180. For James:

    The Higher Power of Religiosity Over Personality on Political Ideology
    Authors
    Authors and affiliations
    Aleksander KsiazkiewiczEmail authorAmanda Friesen
    1.
    2.
    Original Paper
    First Online: 29 August 2019
    59
    Shares
    108
    Downloads
    Abstract
    Two streams of research, culture war and system justification, have proposed that religious orientations and personality, respectively, play critical roles in political orientations. There has been only limited work integrating these two streams. This integration is now of increased importance given the introduction of behavior-genetic frameworks into our understanding of why people differ politically. Extant research has largely considered the influence of personality as heritable and religiosity as social, but this view needs reconsideration as religiosity is also genetically influenced. Here we integrate these domains and conduct multivariate analyses on twin samples in the U.S. and Australia to identify the relative importance of genetic, environmental, and cultural influences. First, we find that religiosity’s role on political attitudes is more heritable than social. Second, religiosity accounts for more genetic influence on political attitudes than personality. When including religiosity, personality’s influence is greatly reduced. Our results suggest religion scholars and political psychologists are partially correct in their assessment of the “culture wars”—religiosity and ideology are closely linked, but their connection is grounded in genetic predispositions.

    Keywords

    Religion Religiosity Personality Ideology

    • Replies: @dux.ie
  181. Aft says:
    @res

    Top 1% SES is z=2.33
    Assume r=0.5 between SES and IQ: z=1.16 IQ parent
    Notch down a tad for heritability: z ~ 0.90 IQ child

    Look up z = 2.33-0.90 = 1.43 in table (=0.076)

    7.6% should be gifted if parents are top 1%

    Magically close to the fractions above.

    (works the other way too: most top percentile kids are not from the top 1% SES)

    (Caveat: lots of gifted programs test for IQ > 125, allow for multiple testing, friendly psychologists, etc.)

    Extra: Top 16% SES, professional class
    z = 1, IQ z =0.5, kid IQ z ~0.4.

    z = 1 – 0.4 =0.6, or 27% of their kids should be in the top 16% cognitively. 66% of their kids should be cognitively above average. ( z>-0.4)

    • Replies: @res
  182. Aft says:
    @res

    Just the first time I’d heard the idea of actual overall intelligence continuing to grow after 18, though seemingly only for the smart, especially the very smart.

    Strikes me as about right, intelligence peaking at grad school age.

    • Replies: @res
  183. Aft says:
    @res

    I tend to think this study is valid, with caveats:
    – CO2 levels change blood PH. One can adapt to this. Perhaps all the way. (A novice runner can barely mentally function while jogging 100m. An experienced runner often does some of their clearest thinking while running miles at a time…). A one-shot test outside one’s norm is not the same as “being on a submarine” for months, global CO2 increasing, etc. etc.

    Strongest evidence for it is the “novelty”/striatal/white-matter loading of the tests that went down: Crisis Response, Strategy, Information Usage, etc. With little detriment on simple or over-learned activities. Matches the literature on pollution, oxidative stress, vascular dysfunctional and mental performance.

    (Oxygen is 20% of air. It wouldn’t decrease materially.)

    • Replies: @res
  184. res says:
    @Aft

    Thanks! Alternatively, a slightly more direct approach is to use the r=0.37 from the paper I linked (child intelligence with parental SES in adolescence) giving z = 2.33 * 0.37 = 0.86 which is essentially the same as your z = ~0.90.

    works the other way too: most top percentile kids are not from the top 1% SES

    “Most” makes sense intuitively, but I had never run the numbers to see that it was >90%.

    I generally prefer simulations to the correlation/z score approach because I like getting a look at the full distributions, but the r/z approach is powerful and quick (i.e. I should use it more often). Pumpkin person uses that approach frequently.

    P.S. This article about gifted education might be of interest here: https://www.theatlantic.com/ideas/archive/2019/10/gifted-and-talented-programs-arent-problem/599752/

  185. res says:
    @Aft

    Good point. I had not really focused on the 18-25 part of the curves before. The trend between absolute level and amount of improvement at later ages seems clear and consistent. The trajectory from ages 10 to 25 (roughly flat or even down) for the -2 and -3 SD lines is rather shocking on reflection.

    P.S. Image reproduced after the MORE to make it easy to see what I mean.

    [MORE]

  186. res says:
    @Aft

    CO2 levels change blood PH.

    That is an incredibly interesting point in this context. People differ in their “normal” blood pH. Does that mean the optimal concentration of atmospheric CO2 would vary depending on someone’s blood pH (and how it relates to “optimal” blood pH)?

    This ties in with metabolic typing. For example, this book: http://www.bloodph.com/
    uses differing breath hold time responses to a nutritional stimulus to assess someone’s “normal” set point.

    Strongest evidence for it is the “novelty”/striatal/white-matter loading of the tests that went down: Crisis Response, Strategy, Information Usage, etc. With little detriment on simple or over-learned activities. Matches the literature on pollution, oxidative stress, vascular dysfunctional and mental performance.

    Could you elaborate on this? Is this evidence for the blood pH effect existing, being the primary effect present, or ? Or are you talking about the overall observation that increasing CO2 impairs intellectual function?

    All of this also raises the questions of whether reducing CO2 concentrations below normal would improve function or whether increasing O2 partial pressure would improve function.

    And how persistent all of these effects would be given adaptation. And as I noted above, is there systematic variation in how these conditions impact different individuals.

    (Oxygen is 20% of air. It wouldn’t decrease materially.)

    I understand that (21% to be more precise, as well as understanding that respiratory O2 and CO2 stoichiometry is roughly 1). But I still think they should have measured O2 for completeness (they measured dozens of distinct VOCs). For example:

    An article about slowly changing O2 concentration in the atmosphere: http://www.columbia.edu/cu/21stC/issue-2.1/broecker.htm

    And Biosphere 2 running low on oxygen:
    https://www.geosociety.org/gsatoday/archive/6/7/pdf/i1052-5173-6-7-sci.pdf

    • Replies: @Aft
  187. Aft says:
    @CanSpeccy

    Not sure if you had a traumatic IQ test in your childhood or what’s going on here. Some posts are lucid, on-point, and deeply researched, and others are careless strawman-lashing.

    No one thinks IQ is everything. It’s just ~50% of variance in cognitive ability. There’s a reason there are math/verbal splits, WAIS subscores, different assessments for different schools, etc.

    Most people don’t even know theirs (though at elite schools the rankings of ability usually become obvious quickly, even without IQ testing. And again, people also pick up on their relative splits.)

    And yes, lots of scammers and grifters make up their supposed IQs.

    IQ is helpful for gifted programs. Those matter.

    Age 12-13 college admissions testing is very helpful (helped me opt out of math classes permanently, and realize I could coast through school focusing on actual solo learning and activities and then ceiling the tests later.)

    College admissions tests matter.

    The world would be a better place (more productive, more meritocratic) with even more IQ testing, cognitive sorting, standardized exams without artificially low ceilings, etc.

    cognitive capacity are not constant but subject to large variation according to immediate environmental factors, and (b) that intelligence is not unitary but involves multiple capacities that vary independently under the influence of environmental stress.

    This is basically true. I didn’t know this was true until I traveled extensively in polluted cities in Asia, but it’s definitely true (see my other comment about specific brain areas).

    Luckily the affected domains don’t really apply to most standardized tests, and rooms tend to have ventilation.

    Whatever happened to you and your IQ test or gifted program envy, the literature is helpful. The concept is helpful. Properly characterized, using Jensen as a canonical source, it’s some of the most sure science on the human body or mind that exists.

    Are there issues here and there? E.g. Jensen/Rushton using Raven’s as a pure g measure when it has a large unique variance–yes, there are.

    Does g/IQ measure things that are lower-order factors? Answers itself.

    Does g/IQ measure personality and other conditions? Again, answers itself.

    Who exactly are you trying to convince of what?

    • Replies: @CanSpeccy
  188. Aft says:
    @res

    pH

    People adapt. Same as they adapt to heat, elevation, intermittent fasting, exercise, and most other stressors.

    A one-shot test shows little about CO2 effects in offices/classroms you work in everyday or atmospheric CO2.

    So the tests feel plausible but not necessarily translatable to broader questions.

    Strongest evidence for it is the “novelty”/striatal/white-matter loading of the tests that went down: Crisis Response, Strategy, Information Usage, etc. With little detriment on simple or over-learned activities. Matches the literature on pollution, oxidative stress, vascular dysfunctional and mental performance.

    Could you elaborate on this?

    There’s a decent literature developing on PM2.5 and ultrafine pollution. Those brain areas and a few others keep coming up and match my subjective experience as well.

    Complex, whole-brain thinking doesn’t hold up so well (white matter is the hardest hit, so communication weakens between distant regions), but simple or learned tasks (embedded in existing gray matter networks) stay about the same. Those 3 areas (Strategy, Breadth of Information considered, Crisis Response) are things I wouldn’t want to attempt on an Orange or Red day.

    CO2 probably hits the same areas with greater vascularization and metabolic needs.

    As to lowering CO2 or boosting oxygen, I’d say live somewhere with clean air, get an air filter, and if you really care a surgical mask helps if PM2.5 is beyond 4ug/m2 and especially if it’s double digits. Air pollution is an order of magnitude more important than the cognitive effects of those two gases. (At normal elevations…)

    • Replies: @res
  189. res says:
    @Aft

    A one-shot test shows little about CO2 effects in offices/classroms you work in everyday or atmospheric CO2.

    So it sounds like what we really need is a before and after test of changes in building ventilation (say 6 months after the change)?

    I’m still curious if blood pH is the primary factor underlying the changes with CO2. And whether that implies an optimal (rather than lower=better) level.

    As to lowering CO2 or boosting oxygen, I’d say live somewhere with clean air, get an air filter, and if you really care a surgical mask helps if PM2.5 is beyond 4ug/m2 and especially if it’s double digits. Air pollution is an order of magnitude more important than the cognitive effects of those two gases. (At normal elevations…)

    That seems sensible. Probably worth being attentive to indoor air quality as well. I have a PM2.5 and VOC meter to measure that.

  190. CanSpeccy says: • Website
    @Aft

    Whatever happened to you and your IQ test or gifted program envy

    LOL. You’ve just said what every IQ-ist seems to say when they’re losing the argument, namely, “You must have a low IQ.”

    Actually, the only IQ test I took that mattered, was the British 11+ on which I scored well enough be admitted to an excellent grammar school that I attended for two years before being transferred for reason best known to my parents to what may have been a gifted program for all I know. At least it was at a school which currently charges about US$60,000.00 a year, not counting all the clothing, and gear and extras that nearly double the cost.

    At the grammar school I finished my second year first in my class. At boarding school I was also first in my class in my first term, but then drifted downward, spending several years paying rather little attention to other than science classes. The radical decline in my academic performance led to a three-day round of psychometric testing by the Institute of Industrial Psychologists, an outfit owned and run by Sir Cyril Burt, then President of the British Psychological Society (and later accused of questionable research ethics).

    The conclusion reached was that other than brilliance at assembling colored blocks in specified patterns, I wasn’t all that bright, but that if I were to attend university I should read mathematics. That was a joke since having day-dreamed through two years of classes under a teacher with a sing-song voice and the personality of a damp rag, my mathematical career was at an end. However, I did attend university and gain the faculty prize in a non-mathematical subject.

    But perhaps I have some of the right genes for math. Certainly, I am highly numerate and our kid won the faculty prize in math at a major N. American university.

    In any case, the fact is you cannot measure multiple independently variable cognitive capacities that make up intelligence by a single number, or IQ. Consistent with that incontrovertible fact, IQ test scores correlate poorly with achievement whether measured by annual income or the receipt of a Nobel Prize in physics.

    Of course if you create multiple cognitive tests, some of them will prove useful for selecting bus drivers or physicists, or whatever. But actual ability in specific areas of activity, whether it be mathematics, the interpretation of history, or violin playing, will be much more accurately measured by subject-specific testing. That is why to suggest that a symphony orchestra should chose a first violinist, or a physics department chose a specialist in quantum mechanics based on an IQ test would be considered inane.

    And even where there is a highly academic- or career-specific test, personality will hugely impact actual career performance. The brain operates as a universal computing engine, which means that anyone can understand anything — given enough time and relevant experience. Yes, there are dim witted people who will not make any substantial progress with quantum mechanics, but among those with IQ of around 125, Richard Feynman’s reported score, a combination of obsessive interest and work may get them a remarkably long way.

    • Replies: @res
    , @Aft
  191. res says:
    @CanSpeccy

    The radical decline in my academic performance led to a three-day round of psychometric testing by the Institute of Industrial Psychologists, an outfit owned and run by Sir Cyril Burt, then President of the British Psychological Society (and later accused of questionable research ethics).

    The conclusion reached was that other than brilliance at assembling colored blocks in specified patterns, I wasn’t all that bright, but that if I were to attend university I should read mathematics. That was a joke since having day-dreamed through two years of classes under a teacher with a sing-song voice and the personality of a damp rag, my mathematical career was at an end. However, I did attend university and gain the faculty prize in a non-mathematical subject.

    Thanks. At least I understand now why you hate IQ testing and researchers so much. Any (specific) idea why the mismatch between the test results and your self perceived abilities?

    It seems like this (unquoted by you) part of Aft’s comment was rather on target.

    Not sure if you had a traumatic IQ test in your childhood

    It is characteristic of you that you chose to quote the less on target part of his speculation.

    Do you think the test did any harm to you other than your ego? It sounds to me like your mathematics teacher was probably a bigger negative influence.

    • Replies: @Aft
    , @CanSpeccy
  192. Aft says:
    @CanSpeccy

    The brain operates as a universal computing engine, which means that anyone can understand anything — given enough time and relevant experience.

    You seem to be making every side of every argument–first it’s that intelligence is not unitary, there are lots of different modules, (all true) thus could never be one number.

    Now you say the brain is a “universal computing engine”: not at all true. (And would support there being “just one number”)

    IQ test scores correlate poorly with achievement whether measured by annual income or the receipt of a Nobel Prize in physics.

    Personality matters for *which field one chooses* but no double digit IQ people are brilliant scientists. Not a single one, ever. And res has been the one posting data on IQ tilts and career paths while you just pretend tests are silly.

    A correlation of ~0.5 (lifetime income without range restriction) or even 0.3 (annual income) are huge correlations for one single number from a few hours of testing. I’m pretty sure you could throw in a 15-minute quiz on attitudes toward money, extraversion, and conscientiousness and derive “a single number” that correlates 0.6 maybe even 0.7 with lifetime income.

    Your complaint is not with g / IQ but with one psychologist. Who showed you’re pretty good with the blocks and math: and then you went on to … be good at exactly those things (patents and science achievement and being “highly numerate”).

    Plus most “IQ” tests aren’t even well-powered or remotely designed to detect optimal school subjects.

    You’re acting like a child who can’t see one slightly scary doctor’s visit doesn’t make medicine an evil pseudoscience.

    However it was used, communicated to you, whatever, grow up and get over it and realize 99% of the time, psychometric testing is helping gifted kids escape the double-digit-IQ-drudgery and psychometricians basically invented all the tools that allow valid achievement tests to be designed for overall assessments and even specific-subject assessments.

    • Replies: @res
    , @CanSpeccy
  193. Aft says:
    @res

    The trend is that people project onto those who predicted their limitations the blame for them having those limitations. Or blame parents or society or anything else they can reach for. It takes some time to see that those are rarely the cause of what was already genetically predetermined. The anger will fade when he realizes the psychologist didn’t cause any harm, the long-term trajectory was always the same.

    It’s clear that the test was probably right: raw talent at math and perhaps extreme talent at spatial / assembly tasks but a total non-fit for something like publishing research at a university. (Inability to engage directly with others here on the merits of arguments also shows this.)

    As to psychologists, and telling kids or even adults things, they have a tendency to not fully grasp the instruments they’re using and be a bit abrasive. (I’ve had experiences with them personally and having them speak to my employees about personality inventories and moved away from that very quickly. So I empathize with the dislike of their at-times poor interpretations or communications. Though the assessments themselves are still extraordinary in their descriptive and predictive power.)

    Also I’d say both parts were pretty much dead-on 😉 (he is protesting a low score after all)

    • Replies: @res
    , @CanSpeccy
  194. CanSpeccy says: • Website
    @res

    At least I understand now why you hate IQ testing and researchers so much.

    Actually, you don’t.

    First, because I don’t hate IQ testing.

    Go ahead test all you want. Just understand (a) that intelligence is not a unitary phenomenon — that the gifts that Mozart possessed were not those of Michael Angelo, nor was Michael Angelo’s genius the same as Newton’s, and (b) that because intelligence is genetically based (as is every feature of the organism) is it vastly modified by environment.

    Any (specific) idea why the mismatch between the test results and your self perceived abilities?

    What the fuck are talking about? You’re just being an idiot troll, as stupid as Mikemikev. As I said, you have bought into a corrupt profession and you demonstrate by your dishonest debating tactics that you have already been corrupted by it.

    • Replies: @res
  195. res says:
    @Aft

    I’m pretty sure you could throw in a 15-minute quiz on attitudes toward money, extraversion, and conscientiousness and derive “a single number” that correlates 0.6 maybe even 0.7 with lifetime income.

    I tend to agree. And find it interesting that apparently no one has ever tried this. Perhaps someone could do this with something like an SMPY/Duke TIP cohort? (even after the fact in a future year followup would be of interest, though less useful for establishing predictive validity) Ideally they could do it with everyone taking the initial test (if you are already taking the SAT how much more effort would this be?), but I think even just testing the passing cohort would be of value.

    • Replies: @CanSpeccy
  196. res says:
    @CanSpeccy

    You pretty clearly implied the test did not capture your abilities correctly. I wish I understood just which nerve I touched with that. Was it my use of “self perceived”? If so, I apologize. I would be fine with something like “demonstrated by future success.”

    Just understand (a) that intelligence is not a unitary phenomenon

    I think I have demonstrated here that I do understand this. Perhaps better than you because I recognize both the unitary and independent aspects.

    (b) that because intelligence is genetically based (as is every feature of the organism) is it vastly modified by environment.

    Vastly is an extremely vague word (and your sentence syntax is garbled, but I think I get the meaning). But given a heritability of ~0.7 (and measurement error is part of the other ~0.3) I think it is hard to justify “vastly” even using a weak definition of it.

    • Replies: @CanSpeccy
  197. CanSpeccy says: • Website
    @Aft

    You seem to be making every side of every argument–first it’s that intelligence is not unitary, there are lots of different modules, (all true) thus could never be one number.

    Glad we agree on that. A conclusion confirmed by the low correlation (averaging less than 0.4) among cognitive traits measured by psychometricians.

    Now you say the brain is a “universal computing engine”: not at all true. (And would support there being “just one number”)

    Well, I’m not going to engage in a “tis” “tisn’t” exchange, but a universal computing mechanism is easy enough to envisage and the computational capacity of the brain can clearly simulate such a mechanism. But that does not meanthat would “support “just one number.”” There is a great deal more to intelligence than mere computation.

    but no double digit IQ people are brilliant scientists.

    LOL. So now the huge value of IQ tests is to discriminate between the double digits and the rest. Look, any school teacher knows which kid can understand a logical argument and which kid is inherently dopey. You don’t need no white-coated psycho to undertake tests to confirm the diagnosis.

    However, I applaud your comment inasmuch as it confirms that IQ tests are of minimal value, allowing only the broadest inferences about intellectual potential in particular areas of achievement.

    However it was used, communicated to you …

    Like Res, and so many other IQ-ists, you resort to the crude ad hominem: i.e., you question the validity of IQ as the measure of intelligence because you got a poor score on an IQ test. But as I’ve repeatedly stated, I have no idea what my IQ test score is or was, no such measure having been communicated to me. Further, I have no objection whatever to psychometric testing, i.e., testing for specific aptitudes such as fitting colored blocks to specific patterns. Apparently such tests can be useful, for example in selecting low accident risks among applicants for driving jobs with a bus company — or so I recall H.J. Eysenck reporting.

  198. CanSpeccy says: • Website
    @res

    I’m pretty sure you could throw in a 15-minute quiz on attitudes toward money, extraversion, and conscientiousness and derive “a single number” that correlates 0.6 maybe even 0.7 with lifetime income.

    That concedes the point, doesn’t it, that IQ correlates poorly with lifetime income.

    • Replies: @res
  199. res says:
    @Aft

    Also I’d say both parts were pretty much dead-on 😉 (he is protesting a low score after all)

    I think he would argue (and in my estimate be correct in doing so) that his score (any of the subtests) was not low. Just lower than one of: 1. his expected (aka “true”) score or 2. his other scores.

    Despite his not showing it very well in much of this thread, CanSpeccy is clearly a smart and knowledgeable man based on the totality of his comments (and enough of his asserted accomplishments have been both consistent with his persona here and verifiable for me to take them at face value, no details because I don’t want to dox anyone) in this blog. We all have our blind/weak spots and hobby horses. I suspect many of mine are glaringly obvious to anyone who reads my comments closely. And I am more guilty than most of lashing out intemperately so am fairly sympathetic even with that.

    What I tend to care about is signal to noise ratio (SNR). If someone has enough signal I will put up with a great deal of noise. Just don’t expect me not to fire back if attacked ; )

  200. CanSpeccy says: • Website
    @res

    You pretty clearly implied the test did not capture your abilities correctly.

    I did not say that nor did I imply it. I said that whoever wrote the report was in error to advise that I should read mathematics at university. The fact that I might well have pursued a mathematical career had my education in mathematics not gone off the rails a short while before is beside the point.

    Parenthetically, it is interesting that those with the potential to do math can easily be put off the subject by difficulty with particular teachers or particular minor problems that create road-blocks. In view of my own experience, I made some effort to aid one of our children who seemed particularly prone to difficulty with math. At pre-school, unable to remember the number seven, for example. At Grade 6 telling his mother he could no longer do math. Ultimately, a teacher gave him a book to help with a particular problem. After that he always read the book and had no further difficulty, going on by reading books, to work through the university math curriculum before completing high school.

    Vastly is an extremely vague word (and your sentence syntax is garbled, but I think I get the meaning). But given a heritability of ~0.7 (and measurement error is part of the other ~0.3) I think it is hard to justify “vastly” even using a weak definition of it.

    Trouble you have is in understanding what you are talking about.

    Measured heritability depends on the environment. If you have a highly homogeneous culture created through near universal exposure to the same media, advertising, entertainment and low-grade educational system, yes, heritability will be high. But that does not alter the fact that the kid who’s raised in a non-typical environment may express genetic potentiality very differently to the average kid.

    That’s why rich people send their children to Eton and Oxford or the US or other equivalents: to get more bang for their genetic buck.

    Conversely, those brought up under adverse conditions, say a kid raised in the Central African Republic (average life expectancy 53 years) or raised manacled in a cage like Tsar Ivan IV, will get a hell of a lot less intellectual bang for their genetic buck.

    • Replies: @res
  201. CanSpeccy says: • Website
    @Aft

    It’s clear that the test was probably right: raw talent at math and perhaps extreme talent at spatial / assembly tasks but a total non-fit for something like publishing research at a university.

    LOL

    My last peer-reviewed journal article, published in 1990, four years after I had quit academa for a career in business, has been cited in the literature close to 500 times.

    As for the bit about “at a university,” here’s the reality. When I submitted that last paper for publication I held three academic appointments, two adjunct, including a full professorship, and one tenure-track, the last at what is rated by the Times Higher Ed. as among the world’s top 30 universities. Of the others, one is rated in the World’s top 40, and the other is my home-town university.

    I don’t really like publishing my cv here but if you insist on inferring bullshit about me, I will put you right.

  202. res says:
    @CanSpeccy

    Not when the sentence preceding that quote was:

    A correlation of ~0.5 (lifetime income without range restriction) or even 0.3 (annual income) are huge correlations for one single number from a few hours of testing.

    Just because it is possible to use more information to improve on predictions based only on IQ does not mean IQ is a poor predictor.

    Perhaps you can give examples of some good predictors you would prefer to use instead of IQ?

    • Replies: @CanSpeccy
  203. res says:
    @CanSpeccy

    You pretty clearly implied the test did not capture your abilities correctly.

    I did not say that nor did I imply it. I said that whoever wrote the report was in error to advise that I should read mathematics at university.

    I took it that the report writer’s error was prompted by unrepresentative test results. If not, why did they give you that recommendation? I will also note, here is exactly what you wrote (emphasis mine): “The conclusion reached was that other than brilliance at assembling colored blocks in specified patterns, I wasn’t all that bright

    Since you seem to think you are bright in things other than block design (and I happen to agree) then I think my “implied” assertion is justified. And, frankly, your denial mystifying.

    The fact that I might well have pursued a mathematical career had my education in mathematics not gone off the rails a short while before is beside the point.

    Less relevant, perhaps. Irrelevant, I don’t think so. You seem to be awfully hard on the psychologist relative to the math teacher given the significance of their respective roles in your life.

    Parenthetically, it is interesting that those with the potential to do math can easily be put off the subject by difficulty with particular teachers or particular minor problems that create road-blocks. In view of my own experience, I made some effort to aid one of our children who seemed particularly prone to difficulty with math. At pre-school, unable to remember the number seven, for example. At Grade 6 telling his mother he could no longer do math. Ultimately, a teacher gave him a book to help with a particular problem. After that he always read the book and had no further difficulty, going on by reading books, to work through the university math curriculum before completing high school.

    This is the kind of conversation I find interesting. And I agree. Did his standardized test results align more with his ability or his school performance when he was having trouble? One of the reasons I am positive towards standardized tests is I think they offer one of the best ways to detect cases like this. Of course then one (here, ideally, it would be the school, but that tends not to be reality) needs to take action. Not everyone has a father like you (or a mother like your wife).

    One thing I struggle with is I learn well by reading. So I tend to just point people at appropriate references. This does not work well for everyone, and some are actively put off and/or offended by my approach.

    FWIW, I had an elementary school teacher who almost put me off school completely. Fortunately, the next year I had one of the best teachers I have ever had for encouraging students in general and me in particular.

    Trouble you have is in understanding what you are talking about.

    Measured heritability depends on the environment. If you have a highly homogeneous culture created through near universal exposure to the same media, advertising, entertainment and low-grade educational system, yes, heritability will be high. But that does not alter the fact that the kid who’s raised in a non-typical environment may express genetic potentiality very differently to the average kid.

    That’s why rich people send their children to Eton and Oxford or the US or other equivalents: to get more bang for their genetic buck.

    Conversely, those brought up under adverse conditions, say a kid raised in the Central African Republic (average life expectancy 53 years) or raised manacled in a cage like Tsar Ivan IV, will get a hell of a lot less intellectual bang for their genetic buck.

    I understand this much better than you seem to think. And I repeat it in full because the points are worth emphasizing. But I stand by my “vastly” critique. Here is what you wrote earlier (and it might be helpful if you ungarbled it so I can be sure I am interpreting it correctly):

    (b) that because intelligence is genetically based (as is every feature of the organism) is it vastly modified by environment.

    What I particularly object to is the “is vastly modified” part. If you had written “can be vastly modified” I would agree while still wanting the caveat that it is much easier to modify IQ downwards by an extremely bad environment than to modify IQ upwards by an extremely good (compared to western norms) environment.

    The key point of the heritability numbers is to make clear how much (or little) effect environment has within the range of normal variation in our society.

  204. CanSpeccy says: • Website
    @res

    A correlation of ~0.5 (lifetime income without range restriction) or even 0.3 (annual income) are huge correlations for one single number from a few hours of testing.

    Huge.
    Huge?
    HUGE?
    Who are you kidding? A correlation of point five means your predictor variable accounts for 25% of the observed variation in what was predicted. Zero point three means your predictor variable accounts for less than 10% of the variation in what was predicted.

    Perhaps you can give examples of some good predictors you would prefer to use instead of IQ?

    I wouldn’t be interested in such a predictor. Neither are you except as proof that IQ predicts something and therefore must mean something. But what it means you don’t know because you don’t know, in any particular case, what environmental and what genetic factors go into it.

    • Replies: @res
  205. res says:
    @CanSpeccy

    I did not write the “huge” sentence. But if you can’t see how stunning it is that a few hours of testing in childhood can predict 25% of variation in lifetime income then I’m not sure what to say.

    I wouldn’t be interested in such a predictor.

    I am surprised you are so intellectually incurious. Or perhaps it would be more accurate to say you realize there isn’t anything close to being as good?

    So how do you want to do things like screening for academic opportunities and jobs? As I have said to you before, you can’t interview everyone.

    Neither are you except as proof that IQ predicts something and therefore must mean something.

    On the contrary, I am quite interested in good predictors. (I am also not a big fan of people who so confidently assert they know what I think. Especially when based on so little evidence. And most of all, as in this case, when simply wrong.) One example is the marshmallow test. This is the reason I responded positively to Aft’s comment about the potential predictive power of coupling some additional tests to an IQ test. My pet idea is that a combination IQ and marshmallow test would be the best two variable predictor we could find. I would love to see someone try to test that hypothesis.

    And if by some chance you did come up with a better predictor than IQ (not holding my breath) I would be very interested in finding out more.

    But what it means you don’t know because you don’t know, in any particular case, what environmental and what genetic factors go into it.

    It is quite clear you have never done any predictive modeling. Nor do you appear to have much interest in it.

    But despite that, there are few better ways to cut through BS and get to the core of what works (and matters) in practice than making models and testing them against reality. (thinking about it, perhaps that ability to cut through BS is why you don’t like it?)

    Presumably at some point we will have relatively good PGS (within the limits imposed by heritability of course) for IQ and we can take a closer look at how relatively important IQ genetic and environmental contributions are for other outcomes.

  206. CanSpeccy says: • Website

    I am surprised you are so intellectually incurious. Or perhaps it would be more accurate to say you realize there isn’t anything close to being as good?

    Why begin with an insolent remark and then immediately take it back.

    As to nothing being as good as an IQ test, what is it good for? Predicting life-time income, which it does very badly? And it’s not surprising that it predicts life-time income badly (why the Hell would anyone need to know that anyway?), since if it fails to measure at least some form of intelligence it would hardly work well, and if it does measure a higher form of intelligence it wouldn’t work very well either, since getting as much money as possible is not what most intelligent people strive to do even in this degenerate age.

    Certainly, I never thought about maximizing income. I was always able to earn a good living so why would I put in 80 hours a week to get more money? Instead I put in 80 hour weeks doing research working for multiple governments as a consultant and as a civil servant, and also in academia. That is, until I ran out of appealing ideas to pursue, whereupon I launched a business, intending to provide myself with a modest pension. As it turned out, I did better in the publishing world than Ron Unz, who has expended plenty of red ink, but has never learned to edit properly. I on the contrary made both a living and a decent pension by selling out to a multi-national.

    As I’ve pointed out before, the only use of an IQ test that I can see is to differentiate among a mass of people when you have to decide in short order who will dig the latrines and who will be trained as an officer. A rough and ready test that reflects both native wit and education, which is what an IQ test is, must be somewhat useful for that.

    But even for that, the usefulness is limited. When my father was inducted into the Royal Air Force in 1941 they gave him not an IQ test but a math test, which revealed not only whether he had the wit to do the math but whether he new how to do it, which is more important when you’re plotting a route around enemy anti-aircraft defenses.

    • Replies: @res
  207. res says:
    @CanSpeccy

    I am surprised you are so intellectually incurious. Or perhaps it would be more accurate to say you realize there isn’t anything close to being as good?

    Why begin with an insolent remark and then immediately take it back.

    Insolent remarks are plentiful in this thread. If you don’t want me to make them then stop doing it yourself. Perhaps you better understand obscenities and simplistic ad hominems? That is what you seem to like using. Perhaps you would prefer me doing the same?

    As for why I chose that structure, it was meant to emphasize the way you seemingly pretend not to be interested in things which are contrary to your beliefs. If you considered the second part “taking it back” you need to work on your reading comprehension.

    As to nothing being as good as an IQ test, what is it good for?

    Sigh. We have had MANY conversations about this. The most notable examples are for screening and placement in academic and military settings. We would probably see more of it in corporate settings as well if not for the infinite wisdom of the US Supreme Court.

    since getting as much money as possible is not what most intelligent people strive to do even in this degenerate age.

    This is a good point. And part of what makes it so impressive that IQ can predict even 25% of the variance.

    But even for that, the usefulness is limited. When my father was inducted into the Royal Air Force in 1941 they gave him not an IQ test but a math test, which revealed not only whether he had the wit to do the math but whether he new how to do it, which is more important when you’re plotting a route around enemy anti-aircraft defenses.

    Yes. We have also talked about the utility of finer grained testing for finer grained placement. One question though. Do you think they gave that math test to everyone? My bet would be they used a general test (or the math portion of it) as a pre-screen to choose people to give the AA plotting version of the math test. Or maybe they just gave a harder math test to divide the people most able in math between areas like cryptography, AA plotting, operations research (logistics is important! https://newsroom.kpmg.com.au/amateurs-talk-strategy-professionals-talk-logistics/ ), radar data processing, etc.

    I did better in the publishing world than Ron Unz

    How did your readership compare?

    Also, you spent an entire paragraph talking about how money is not important then contradict yourself in the final sentence by using it as the metric you consider important. And do you really want to get in a financial pissing contest with Ron? Do you not understand that his emphasis in publishing is to influence opinions rather than that base money grubbing you so deplore?

    You really do think well of yourself. How about putting some of that ability to work doing the following?
    – Making an effort to understand how multiple things can vary both together and separately.
    – Making better arguments.
    – Responding to my good arguments in addition to my insolent comments.
    – Cutting down on careless mistakes like “he new how to do it”.

    • Replies: @CanSpeccy
  208. CanSpeccy says: • Website
    @res

    The trouble with you, Res, is that you are not a scientist. You are a careerist who has joined the IQ-ist racket. And for what? A career selling bullshit.

    If you had a scientific mentality, you would ask yourself some serious questions about intelligence. For example, you would ask what are the many ways in which information is acquired and used, and how might those processes be measured.

    You might also ask yourself the reasons, whether scientific or practical, for measuring those variables. Or do your repeated references to life-time income answer that question?

    But whether your basic idea is that intelligence is a measure of the money a person gets, your most recent claim that IQ predicts 25% of the variation in life-time income is refuted by Zagorsky’s data, which have already been discussed here:

    https://miro.medium.com/max/1158/1*Txc8Deu_EEM7pH7ZL0k4cg.png

    Those data indicate that the predictive value of IQ is trivial: less than ten percent. Moreover, according to the New Scientist:

    On the surface, Zagorsky’s analysis confirms the findings of previous studies linking higher intelligence with higher income. … people with higher intelligence scores also had greater wealth. … But when Zagorsky controlled for other factors – such as divorce, years spent in school, type of work and inheritance – he found no link between IQ and net worth.

    But still you cleave to the IQ-ists obviously false claim that intelligence is a single indivisible quality that can be measured by a single number, i.e., on a uni-dimensional scale.

    To anyone with the capacity to think, that is obvious nonsense. The abilities of individuals differ qualitatively and to a massive degree. The IQ-ist’s claim that the existence of the common factor g proves otherwise, is false. The value of g reflects the degree of correlation among the small number of intellectual capacities covered by an IQ test. But those correlations mostly trivial, as this correlation matrix confirms:

    https://3.bp.blogspot.com/-4_cOTm5OG8M/WP4suocVPWI/AAAAAAAAEWM/FSJg8nXxR5Uh4JSrGPLVia3ytPRC4L0swCLcB/s1600/Correlation%2BMatrix%2BMental%2BAbilities.png

    But in defense of a nonsensical thesis you, AFT, and other defenders of the IQ-ist fraud ridicule anyone who questions it by calling them low IQ. What’s that if not proof of charlatanry. You and AFT presume to rate my intelligence without even an IQ test. LOL.

    • Replies: @CanSpeccy
    , @res
  209. CanSpeccy says: • Website
    @CanSpeccy

    But, to be more positive, why not abandon you IQ-ist will o’ the wisp, or cognos fatuus, and start thinking seriously about the huge opportunities that the study of intelligence provides.

    Memory for example, How to quantify it?

    A google scholar search seems to indicate that there is not one scholarly journal article reporting an attempt to quantify musical memory.

    So there you are, just one of many fields in intelligence research wide open.

    In particular, how is it that some people can remember complete sonata or concerto or even an opera at a single hearing?

    Even my six-year-old grand-daughter can remember a song at a single hearing. What’s going on here? How do you quantify it?

    Surely a great opportunity here for collaboration, one would think, among psychologists information theorists, and neurologists.

    Think it over. You might come up with dozens of questions at least as important, and at least as compelling in demonstrating the vast complexity of intelligence, a general term for the functional properties of the human brain, the most complex known object in the universe.

    • Replies: @res
    , @James Thompson
  210. res says:
    @CanSpeccy

    The trouble with you, Res, is that you are not a scientist. You are a careerist who has joined the IQ-ist racket. And for what? A career selling bullshit.

    Do you think I am a psychologist working on IQ? Despite how many times I have said my background is engineering?

    I enjoy all of the attempts to analyze me and my motivations. Because when they are wrong (which seems to almost always be the case, which, to be honest, I find kind of frustrating) I get a clear look at people’s analytical skills and judgment.

    You might also ask yourself the reasons, whether scientific or practical, for measuring those variables. Or do your repeated references to life-time income answer that question?

    We have talked about this before. It could actually be an interesting conversation if you would engage in a way other than just disparaging everything you don’t like. The primary uses I see for measuring IQ (or similar) is for screening (which I have mentioned enough times by now you wouldn’t have to ask if you were an honest interlocuter) and for use as an explanatory variable. For example, helping people understand that “disparate impact” can often be caused by simple differences in ability. Or to get a better sense of which countries might have difficulty succeeding because of a lack of intellectual capital in the population.

    your most recent claim that IQ predicts 25% of the variation in life-time income is refuted by Zagorsky’s data, which have already been discussed here:

    So the scatterplot which does not even show all of the relevant data again. If anyone else wants to revisit the discussions about Zagorsky’s work we had recently see these threads:
    http://www.unz.com/jthompson/the-intelligent-investor/
    http://www.unz.com/jthompson/the-wages-of-intellect-part-2/

    Regarding your New Scientist quote. Yes, when you correct for variables highly correlated with IQ (e.g. years of education) the connection with IQ decreases. Shocking. Unless you understand what happens when you do regression with correlated explanatory variables.

    But still you cleave to the IQ-ists obviously false claim that intelligence is a single indivisible quality that can be measured by a single number, i.e., on a uni-dimensional scale.

    One of the best ways of seeing that someone has a weak case (or is simply bad at debate) is that they have to resort to a strawman argument. This has been adequately covered above. Just search the thread for “strawman”.

    But in defense of a nonsensical thesis you, AFT, and other defenders of the IQ-ist fraud ridicule anyone who questions it by calling them low IQ.

    Please show me where I have done that. I have in fact done the opposite with you repeatedly. I can live with your ad hominems and strawmen (pathetic as they are), but the lying (see PS as well) really is intolerable. And if you had any intellectual honesty you would realize that yourself.

    P.S. Are you going to own up at some point to how your denial in comment 203:

    You pretty clearly implied the test did not capture your abilities correctly.

    I did not say that nor did I imply it.

    was refuted by my comment 206 or are you just hoping if you ignore that long enough it will go away? Here is the relevant excerpt from comment 206:

    here is exactly what you wrote (emphasis mine): “The conclusion reached was that other than brilliance at assembling colored blocks in specified patterns, I wasn’t all that bright

    Since you seem to think you are bright in things other than block design (and I happen to agree) then I think my “implied” assertion is justified. And, frankly, your denial mystifying.

    Rarely does one get to see a statement so decisively refuted. Thanks for the opportunity.

  211. res says:
    @CanSpeccy

    That’s pretty funny coming from someone who in comment 176 complained:

    Fine diversion by irrelevance.

    To actually give a serious response though.

    1. My musical ear and talent are not strong points of mine. I think it is sensible to focus on something which more closely matches my skills.

    2. The IQ field seems like quite a rich enough vein to me. Especially given the combination of a large body of prior research combined with all of the people in the Current Year who stick their fingers in their ears and pretend that makes the research not exist.

    3. The ongoing tie in with genetics work is a sweet sport for new knowledge as well.

    • Replies: @CanSpeccy
  212. CanSpeccy says: • Website
    @res

    That’s pretty funny coming from someone who in comment 176 complained:

    Fine diversion by irrelevance.

    There’s nothing diversionary about measuring the capacity for information acquisition. It’s part of what intelligence is, which is — to say it once again in the hope that the fact might at last sink in — the ability to acquire and use information.

    1. My musical ear and talent are not strong points of mine. I think it is sensible to focus on something which more closely matches my skills.

    I don’t see what your talents of lack thereof have to do with it. The scientific questions relate to how much information there is in a musical composition, how that information is encoded by the brain, in what part of the brain is it encoded, how much is retained, and if we have an obsession about making comparisons among individuals, how much individuals vary in the capacity for this type of memory.

    Visual, olfactory, gustatory, proprioceptive, and tactile memory present a similar sets of challenges, although certainly the way those various types of information are processed and the parts of the brain involved will be different.

    The IQ field seems like quite a rich enough vein to me.

    My God, you really are hopelessly stuck in a futile rut. You think you can characterize the potential of the human brain, the most complex structure in the known universe, with a single two- or three-digit number.

    You’d have made a fine alchemist, or voodoo artist. But a scientist? You seem to lack the curiosity gene altogether. Instead you want to wrap up a bunch of variables with your IQ test score to create a “predictive model” for life-time income. For what reason? So that idiot IQ-ists, people like the liar Jordan Peterson, can go on saying that “IQ science” is the best thing psychology has going for it because it predicts life-time income, except it doesn’t to any useful degree. And indeed, even if it predicted with total accuracy, what would be the value of the prediction? None at all as far as I can see. No, its only value is to the IQ-ists to prove their bunk isn’t bunk, except unfortunately for them, it is bunk and always will be.

    • LOL: res
    • Replies: @mikemikev
  213. dux.ie says:
    @James Thompson

    In Becker’s dataset there are only Flynn effects on Spatial IQ. Is there any large dataset on global spatial IQ? Need them to prove why East Asians have higher Spatial IQ.

    • Replies: @res
  214. dux.ie says:
    @Godfree Roberts

    Angry vs IQ, religiosity and personality

    Recently Gallup has a poll on the global percentage of people from various countries who felt angry the day before. Mashing in with the national average IQ, Big5 personality score and %religious, the %Angry is negatively correlated with national average IQ, significant with religion and only the Neuroticism personality is significant. From the Gallup data most of the countries for %angry are Christian and Muslim countries. Thus Buddhism is not a factor and Islam and Christianity are more or less negatively correlated buffered by the agnostics. So if either one does not show up to be significant they could be masked out by the correlation with the other. The interaction term between Christian and Muslim is also considered to see if one triggers the other but it is not significant.

    As expected Neuroticism (tendency to experience negative emotions, such as anger, anxiety, or depression) has the largest effect size for personality. From the poll the Filipinos seems to be excessively angry. Lebanon, Greece and Bangladesh can remain cool below their estimated scores. Finland and Estonia must be the most serene among the lot. There are no data for HongKong, it could be an outlier with respect to religion.

    [MORE]


    lm(Angry ~ IQdb+Christian*Muslim+B5Extr+B5Agr+B5Cons+B5Neu+B5Ope));

    Parsimonous regression equation with significant terms remaining:
    lm(formula = Angry ~ IQdb + Muslim + B5Neu)

    Coefficients:
    Estimate Std. Error t value Pr(>|t|)
    (Intercept) -5.14000 18.73270 -0.274 0.78499
    IQdb -0.28632 0.09807 -2.920 0.00537 **
    Muslim 0.10667 0.03509 3.040 0.00386 **
    B5Neu 0.95204 0.36526 2.606 0.01222 *

    Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

    Residual standard error: 6.161 on 47 degrees of freedom
    Multiple R-squared: 0.4073, Adjusted R-squared: 0.3694
    F-statistic: 10.76 on 3 and 47 DF, p-value: 1.676e-05

    — Sample results, EstAngry from the model
    Rank|ISO | AngryPoll | EstAngry | Resi | Country
    1 MAR | 41 | 39.19 | 1.81 | Morocco
    2 TUR | 40 | 39.75 | 0.25 | Turkey
    3 PHL | 34 | 19.72 | 14.28 | Philippines <—
    5 ITA | 30 | 22.47 | 7.53 | Italy
    6 IND | 27 | 26.30 | 0.70 | India
    10 ISR | 24 | 24.02 | -0.02 | Israel
    12 LBN | 23 | 38.39 |-15.39 | Lebanon <—
    13 USA | 22 | 24.28 | -2.28 | United States <—
    15 BGD | 21 | 46.32 |-25.32 | Bangladesh
    18 ZAF | 20 | 27.56 | -7.56 | South Africa
    19 FRA | 19 | 18.27 | 0.73 | France
    27 GRC | 16 | 57.20 | -41.2 | Greece <—
    29 DEU | 16 | 16.54 | -0.54 | Germany
    30 KOR | 16 | -6.68 | 22.68 | Korea, South
    35 CAN | 15 | 17.46 | -2.46 | Canada
    36 UKR | 15 |-64.97 | 79.97 | Ukraine <—
    37 JPN | 14 | 13.59 | 0.41 | Japan
    38 GBR | 14 | 12.25 | 1.75 | United Kingdom
    39 CHE | 13 | -0.97 | 13.97 | Switzerland
    42 TWN | 11 | 1.57 | 9.43 | Taiwan
    45 HRV | 10 |-34.27 | 44.27 | Croatia
    48 MEX | 8 | 7.31 | 0.69 | Mexico
    49 NLD | 8 |-16.69 | 24.69 | Netherlands
    50 FIN | 7 | 5.51 | 1.49 | Finland
    51 EST | 6 | -7.04 | 13.04 | Estonia

  215. mikemikev says:
    @CanSpeccy

    My God, you really are hopelessly stuck in a futile rut.

    That happens with some people. I think the type is described well in a classic book:

    …knew nothing at all about yesterday and repeated his same old twaddle as though nothing had happened…

    • Replies: @res
  216. @CanSpeccy

    In particular, how is it that some people can remember complete sonata or concerto or even an opera at a single hearing?

    Even my six-year-old grand-daughter can remember a song at a single hearing. What’s going on here? How do you quantify it?

    Surely a great opportunity here for collaboration, one would think, among psychologists information theorists, and neurologists.

    Carl Emil Seashore’s test of musical ability contains a test for the ability to remember melodies. 1915.

    • Replies: @res
    , @CanSpeccy
  217. res says:
    @James Thompson

    Thanks! Here is an article (well, transcript of an audio presentation) about that:
    https://www.uh.edu/engines/epi1736.htm

    That looks like an interesting series for those who are intellectually curious. This page links to a full list of episodes with titles:
    http://www.uh.edu/engines/

    Here is a 1998 review article:
    Musical Aptitude Testing: From James McKeen Cattell to Carl Emil Seashore
    http://www.public.asu.edu/~aajth/Personal/MusAptTest.PDF

    The article is only 6 pages of text (plus another 6 pages of references!). I wanted to include the abstract here, but the OCR is terrible. The TLDR version is that Cattell included musical discrimination in his 1890s tests of sensory discrimination, but when that failed to predict academic grades most of the research followed Binet and Henri’s work (Seashore being the notable exception).

    I think the full article is worth a read. Some good history (most notably of Cattell and roots of sensory perception testing in a German tradition) and observations about who studied what when and why which connect to some of CanSpeccy’s rants. References 47 and 48 have some more recent followup on the musical work.

  218. res says:
    @mikemikev

    Nailed it.

    I wonder if CanSpeccy realizes just how obvious the projection in his comments is.

    • Replies: @CanSpeccy
  219. CanSpeccy says: • Website
    @James Thompson

    Thanks, James, for the reference to Seashore. A useful intro to a fascinating field of experimental research.

    • Replies: @James Thompson
  220. res says:
    @dux.ie

    In Becker’s dataset there are only Flynn effects on Spatial IQ.

    Could you elaborate on this? I just searched Becker’s V1.3 spreadsheet for “spatial” and all I saw was 23 hits in reference titles.

    Is there any large dataset on global spatial IQ?

    I assume you are aware of the spatial ability data in Lynn (1991)?
    https://www.researchgate.net/publication/247325974_Race_differences_in_intelligence_A_global_perspective
    Only a subset of countries (search the text for “spatial”, the data is scattered through different tables), but better than anything else I know of. Better spatial ability by country references appreciated.

    P.S. Since we are talking about spatial ability, this popular article might be of interest:
    Recognizing Spatial Intelligence
    https://www.scientificamerican.com/article/recognizing-spatial-intel/

    P.P.S. Some interesting but a bit odd speculation about East Asian visuospatial ability: https://www.pienisalaliittotutkimus.com/2018/05/11/why-east-asians-have-iq-profile-skewed-to-visual-spatial/

    • Replies: @dux.ie
  221. @CanSpeccy

    Richard Lynn looked at the literature, and there is some cross-cultural stuff as well. Haven’t got the references to hand, but I think it is this one.

    Sanderson, H. E. (1933). Differences in Musical Ability in Children of Different National and Racial Origin. The Pedagogical Seminary and Journal of Genetic Psychology, 42(1), 100–119. doi:10.1080/08856559.1933.10534231 

    • Replies: @res
  222. CanSpeccy says:
    @res

    Come on Res, don’t be a gutless wimp. If you want to spew insults, at least have the courage to address them directly to your target. And spell it out. What is this projection you ridicule me for? What are the unconscious impulses or fears that I am seeking to deny. Or is it you who are projecting: out of a fear that your entire intellectual construct of IQ-ist nonsense may disintegrate before your eyes?

    • Troll: mikemikev
    • Replies: @res
  223. res says:
    @CanSpeccy

    Come on Res, don’t be a gutless wimp. If you want to spew insults, at least have the courage to address them directly to your target.

    Give me a break. I have responded to plenty of your comments directly. In that particular case I thought the insulting nature of my statement was increased by talking about you in the third person as if you were not present. To my mind a feature, not a bug.

    What is this projection you ridicule me for?

    In this case it was you accusing me of being stuck in a futile rut. When it is quite obvious that the statement is more applicable to your comments here than mine (and my complimentary response directed to mikemikev was prompted by him capturing your behavior so well with his quote). Your persona here is very much that of a one trick pony. It is unfortunate that trick is so tedious.

    Another example is you accusing me of a lack of intellectual curiosity. If anything I have the opposite problem of maintaining focus on the topic at hand rather than haring off after other things I find interesting along the way (I would think that is obvious from my comments here ; ). When I receive criticism which is so obviously off target I think it is safe to assume I am seeing projection.

    Shall I continue?

    Or is it you who are projecting

    Perhaps you would care to give an example where you think I am projecting? I am sure you can find at least one in the almost million words of my Unz Review comments. Especially given the two examples (and counting) I see from you in this thread alone.

    P.S. I am still trying to decide if CanSpeccy’s “gutless wimp” was projection. Because he would have to be pretty deluded to think I shy from confrontation here. What does everyone think?

  224. res says:
    @James Thompson

    Thanks. That paper is available on sci-hub.se
    No abstract, but here are the conclusions.

    VI. CONCLUSIONS
    1. Differences which we consider significant have been found to exist between various racial and nativity groups in the components of musical ability as measured by the Seashore and Kwalwasser-Dykema music tests.
    2. According to our results:
    a. The Jewish group shows a marked superiority to all other groups except the German, which ranks a close second.
    b. The Polish group tends to be markedly inferior in tests of
    musical ability.
    c. The Negro group shows a definite inferiority to all other racial groups except in performance on the test of rhythm discrimination.
    d. The Italian group tends to hold a median position in the five racial groups.
    3. Low intercorrelations were obtained between the Seashore and K-D tests of pitch, intensity, and memory.
    4. The majority of the reliability coefficients were found to be low.

    Too bad there were no Asian groups.

    P.S. Can you imagine that paper being published today?!

    • Replies: @James Thompson
  225. CanSpeccy says: • Website

    We seem to have been about as rude as it is possible to be without resort to elaborate direct personal insults at which I have no special aptitude. Therefore, I will grant you the last insult, i.e., you last comment, and add nothing further to the debate as to whether claiming to measure intelligence with a single number is sensible or not.

    That Mikemikev agrees with you so whole-heartedly seems to me like evidence for my side of the argument, but no doubt you disagree. I note your disagreement and wish you a happy Canadian Thanksgiving.

    • Replies: @res
  226. res says:
    @CanSpeccy

    A happy Canadian Thanksgiving (and American Columbus Day) to you (and the others here) as well.

  227. dux.ie says:
    @res

    >> In Becker’s dataset there are only Flynn effects on Spatial IQ.

    In file N-IQ-DATA.V1.3.2.xlsx, in sheet NAT, column SIQ-FLynn-Effect, Def.: Mean annual changes in Spatial-IQ .

    > Lynn’s Spatial IQ data

    That is interesting but there are too few data points for stats. At least that can be used to back up qualitative assertions.

    Most people are fixed on that IQ mostly depended on thousands of SNPs (plus environemnts). However I found that simply from 2 genetic mutations many observable IQ trends can be explained. They can explain

    – why Ashkenazis on average have high IQ, especially high verbal IQ but moderate on Spatial IQ, but not for the general native Jew in Middle East
    – why East Asians on average at the population level have relatively high IQ, especially Spatial IQ but relatively very poor on Verbal IQ
    – why the mark contrast between the national average IQs between neighbouring countries of India and China
    – the IQs of Ice and Sun people.
    – the IQ difference between East Asians and NatAms even though they might had shared common ancestors 12,000 years ago
    – some other IQ effects on people too easily triggered and so best not to talk about them. The explanations might not be true but the experimentally measured scores are still there.

    For quite some time now I have been pushing that 2 genetic mutations, the alcohol metabolism impairement mutation ADH1B*2 and the immune system mutations the CCR5 Delta32 from just their frequency distributions in various populations or countries statistically correlated to the average national IQ.

    Alcohol progressively damage the brain. The ADH1B*2 mutation highly sensitizes the carriers even to small amount of alcohol which make them physically sick that most of them become teetotalers and thus cumulatively less damage to the brain, but those that drinks will have high chance of getting uncurable esophageal cancer. China alone has about 50% of the global esophageal cancer cases. On average Europeans have zero ADH1B*2 mutation. The effect of the % distribution of ADH1B*2 on IQ is best seen on a chart but the chart hosting service I used had closed down, so you have to imagine that with the regression equation,

    IQdb = +0.143*ADH1B2 +92.7; #n=44; Rsq=0.275; p=0.0002545 *** (VVSig)

    %ADH1B*2 distributions:
    Pop | %ADH1B*2
    EastAsian | 73.9 <—
    Ashkenazi | 27.0 <—
    Hispanic | 5.6
    European | 3.8
    Fin | 0.5
    SouthAsian | 5.0

    East Asians because of the large %ADH1B*2, thus on average at the population level have IQ advantage over the Europeans. Ashkennazis have moderate %ADH1B*2 but significantly higher than than for European. The historical ADH1B*2 mutation wave swapt from Middle and mostly to East Asia bypassing India from about 10,000 years ago. Thus very few ancestors of the Nat Americans crossing the Bering Strait had that mutation, in fact they have the ADH1B*3 mutation which could be worse for IQ. ADH1B*2 (rs1229984T) is in the set of significant SNPs for EA3 even though it was derived only from the European sample (limited single digit % range).

    The mechanisms of CCR5 Delta32 are not well studied but the effects on IQ can be observed at the population level. There are reports of the effect of Delta32 directing the flow of stem cells to the damage sites of the brain. CCR5 Delta32 is best known for HIV resistance and the only case of cured HIV from Delta32 bone marrow transplant.

    CCR5 Delta32 mutation so happen to be originated in Northern Europe and spread from there. Thus it is not cold that enhances IQ but the present of CCR5 Delta32 originated there by chance that Ice people are smarter than Sun people. Another collaborating fact is that Intuits or Siberians from the Artic are not that smart. The ancestors of the Ashkenazis migrated to Northern Europe about 1,000 years ago and pick up the CCR5 Delta32 mutation by admixture and so they are different from the Native Jew in Middle East.

    https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.0030339

    Recently there are refereed reports of CCR5 Delta32 enhancing cognitive performance, especially after brain damages after stroke.

    IQdb = +90.71*Delta32 +86.26; #n=67; Rsq=0.357; p=9.423e-08 *** (VVSig)

    A better fit for the non-linear effects,

    IQdb = +6.698*Log(Delta32) +111.5; #n=65; Rsq=0.469; p=3.062e-10 *** (VVSig)

    %CCR5 Delta32 distributions:
    Pop | %Delta32
    Fin | 13.8 <—
    European | 10.7 <—
    Ashkenazi | 12.1 <—
    EastAsian | 0.0
    SouthAsian | 0.0

    The recent paper by Joy et all in Cell Journal confirmed that CCR5 Delta32 enhances cognitive ability,

    https://www.cell.com/cell/fulltext/S0092-8674(19)30107-2
    “CCR5 Is a Therapeutic Target for Recovery after Stroke and Traumatic Brain Injury”

    In 446 study patients, who had their cognitive assessments available (Figure S6D), there were 68 total (CCR5-D32) carriers (15.2%) (Table S1). 89.7% of the carriers were Ashkenazi in their origin, compared to 57.6% in non-carriers, as would be expected with the genetic association of this mutation. CCR5-D32 carriers were more educated (p < 0.001) but did not differ in cardiovascular risk factors.

    The Joy paper further showed the finer grain effects of CCR5 Delta32 enhances Verbal IQ but worse for Spatial IQ,

    Verbal Function
    MonthAfter | CCR5++ | Delta32
    0 | 89.6 | 85.6
    6 | 90.4 | 92.2
    12 | 89.1 | 95.9

    VisuoSpatial
    MonthAfter | CCR5++ | Delta32
    0 | 97.8 | 101.5
    6 | 96.7 | 97.7
    12 | 99.6 | 98.3

    Thus Ashkenazis with both ADH1B*2 and CCR5 Delta32 mutations have on average higher population IQ than the East Asians, especially on Verabl IQ. The East Asians with higher pop level %ADH1B than the European with %Delta32 and so have some advantage on the pop IQ especially on Spatial IQ (with CCR5++) but the East Asians suffer a double negative whammy (ADH1B*2 and CCR5++) on Verbal IQ.

    With no known large dataset on Spatial IQ, the effects can be explored with GRE Quant and Verb as the proxies,

    GREquant = +8.84*ADH1B2 +154; #n=45; Rsq=0.433; p=8.958e-07 *** (VVSig)
    GREverb = -6.18*ADH1B2 +152; #n=45; Rsq=0.28; p=0.0001874 *** (VVSig) negative effect size

    GREquant = +10.4*Delta32 +154; #n=70; Rsq=0.0339; p=0.1269 (NotSig) no enhancement
    GreVerb = +36*log(Delta32) +148; #n=70; Rsq=0.202; p=9.426e-05 *** (VVSig)

    • Replies: @res
    , @res
  228. @res

    Thanks for the summary. No, I don’t imagine it would be published today, but it would still be interesting if the same results would be found today, given the easy availability of so much music.
    Worth a research project.

    • Agree: res
  229. res says:
    @dux.ie

    In file N-IQ-DATA.V1.3.2.xlsx, in sheet NAT, column SIQ-FLynn-Effect, Def.: Mean annual changes in Spatial-IQ .

    Thanks! That column is also in V1.3 (now that I know what to look for). (BTW, V1.3.3 became available July 1st, I just downloaded it) I also see the definition in the V1.3 and later manuals, but not V1.1.

    Do you know where that data came from? The manual says:
    Method: Taken unaltered

    There are only 8 countries with SIQ-FLynn-Effect so I wonder if the data has the same original source as

    > Lynn’s Spatial IQ data

    That is interesting but there are too few data points for stats. At least that can be used to back up qualitative assertions.

    (BTW, I agree with your assessment. I doubt you will be able to find a large enough spatial dataset for good statistical validity–but we can hope. I wonder if there is any chance of the SMPY researchers advocating for a spatial test to be added to PISA or similar tests.)

    That Lynn paper gives Visuo-spatial IQs of
    US 100
    Australia 104
    Belgium 101
    Britain 102
    Sweden 104
    British Indians 89
    Japan 103-114
    Hong Kong 114
    (some numbers for Asians in the West)
    Nigeria 81
    South Africa 69
    (some numbers for mixed B/W populations)
    (some numbers for Amerindians, interesting because their profile is quite tilted towards spatial)
    Maoris 87

  230. res says:
    @dux.ie

    For quite some time now I have been pushing that 2 genetic mutations, the alcohol metabolism impairement mutation ADH1B*2 and the immune system mutations the CCR5 Delta32 from just their frequency distributions in various populations or countries statistically correlated to the average national IQ.

    I saw some of your posts on that. Very intriguing work. The R^2 seem high, but not out of the realm of coincidence. For example, could some of the apparent effect of those SNPs be caused by their correlation with continental races and not just their direct effect on IQ? I think this concern is especially relevant given what Davide Piffer’s work has shown regarding the predictive value of a small number of SNPs on country IQ. Idea being that those SNP frequencies correspond to overall selection effect on IQ.

    Perhaps as a check you could try running your analysis on other SNPs (starting with the top IQ hits used by Piffer) and see how those compare with ADH1B*2 and CCR5 Delta32 in R^2?

    I suspect you might be onto something given the way your idea speaks to all of your bullet points and also gives plausible causal effect pathways for those SNPs. Have you tried talking to Davide Piffer? I think his work might be synergistic with yours.

    I find your cold effect points especially interesting. But would note the two explanations are not exclusive. Perhaps the cold matters for selection, but the source material (available genetic variation) matters as well? BTW, how do your SNPs fit in with the spatial tilt seen in the Lynn Amerindian data? I seem to recall other comments of yours talking about them.

    some other IQ effects on people too easily triggered and so best not to talk about them.

    If you have any way to refer to them obliquely enough to avoid triggering I would be interested in hearing more.

    Have you looked into this CNV at all? https://en.wikipedia.org/wiki/DUF1220
    There is some speculation it might be an important contributor to intelligence.

    P.S.

    the chart hosting service I used had closed down

    Ouch. So all of the graphs you included in past comments are broken links now?! That is a bummer. I tried looking for one on the internet archive and it looks like they did not archive tinypic.com? Even if the image was included in another page which was archived.

    • Replies: @dux.ie
  231. dux.ie says:
    @res

    > could some of the apparent effect of those SNPs be caused by their correlation with continental races and not just their direct effect on IQ?

    The mutation diffused from the originating site. It is unavoidable that there will be continental effects. The more direct test can be done on lab mice. I avoid mentioning the lab mice experiment which specifically taken an common lab mice ancestral cohorts with CCR5++ and split them randomly into the control group and the experiment group which on average there are no difference between the two groups. The embryos from the experiment group are artificially knocked off the CCR5 gene to create the next generation with CCR5++, CCR5+- and CCR5– variants. Those with CCR5++ can be reassigned to the control group. Those with CCR5+- and CCR5– are selectively bred separately to produce the final experiment group with purely artificially created CCR5– sample, and the cognitive performances of the two groups are compared. So in this case there is no “continental effects” as the two samples are all descendants from the same ancestral cohorts. But there are group of people like to add that “the results are only for lab mice”. So I prefer to talk about the effect of CCR5– on the human neurological recovery from stroke.

    I also keep mentioning that it is “on average at the population levels” as on personal level there could be other particular interacting genes at play but at the population level the effects are averaging out.

    > see how those compare with ADH1B*2 and CCR5 Delta32

    I am more interested in those that might have plausible mechanisms. For example I also kept seeing one particular SNP associated with the Crohn disease popping out with other traits like sleep duration, drinking habits, schizo, etc. I have no idea how to link that with IQ except that it might be significant for the Ashkenazis.

    > refer to them obliquely enough

    Well the South Asians seem to have no major problem with the uncurable esophageal cancer. The one pop seems to exclusively have a different ADH1B*3 mutation as the safety net against alcohol dependence. The exclusivity means that it automatically being associated with the other traits present. Is it the cause? I don’t know. There are too few (4) data points.

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4682545/

    • Replies: @res
  232. res says:
    @dux.ie

    Thanks.

    > see how those compare with ADH1B*2 and CCR5 Delta32

    I am more interested in those that might have plausible mechanisms.

    I understand that (and agree). The point of comparing with the other IQ SNPs is to get an idea of how the R^2 values compare in your models. Which in turn might give some idea of how much of the effect you are seeing is causal (I would expect the truly causal SNPs to have higher R^2). Alternatively, you could try looking at random SNPs to get a sense of what their distribution of R^2 would be in the model.

    Your GREquant and GREverb models are of particular interest for the comparison because they are looking at specific abilities and not just the IQ/EDU information used in the GWAS.

    Also, your Log(Delta32) observation makes me think it is more likely you are picking up selection rather than (or in addition to) direct causal effects.

  233. Factorize says:

    Back to 1500 IQ humans. The article in Nature Genetics from July 2018 included a list of 127 IQ/EA causal variants. If these SNPs were CRISPRed to their cognitively maximal genotype, then population IQ/EA would increase by over 2 SD. I had been unaware that there is already this much potential intelligence enhancement available. Notably, a 2 SD gain would override any existing racial differences in IQ, and as these are causal SNPs they should, one might assume, increase IQ for all humans. (A recent article suggested IQ uplift currently might only provide a gain of ~3 IQ points, yet this was assuming only the use of embryo selection.): How will humanity possibly cope with the profound cognitive transformation that is approaching?

    • Replies: @James Thompson
  234. @Factorize

    You may have already explained this, but how have you calculated this massive improvement?

  235. Factorize says:

    The sum of all the positive betas among the 127 causal SNPs is 0.99 SD. If someone were to receive two copies of each of these positive SNPs, then optimal gain is 1.98 SD. Since the positive and negative effect SNPs are largely symmetric and thus cancel each other out leaving an expected value of 0 SD at the population level, we could simply estimate the gain as ~2 SD for an optimized genotype.

    The actual calculation using: sum (p^^2 * 2* beta + 2* p*q*beta) for the negative and positive effect SNPs separately is: 0.322 and -0.354. Thus, the actual population scale expectation for these SNPs is negative -0.032 SD. The gain in IQ/EA for the population would be greater than 2 SD.

    These SNPs have been determined to be causal for IQ/EA with high probability. The only assumption made here (which is admittedly substantial) is that 127 SNPs could be homozygously CRISPRed into the optimal state. However, given that CRISPR has already been applied in humans to change the germline, suggests that such a large scale modification is not entirely beyond the realm of the possible.

    2 SD IQ uplift would cause extreme social disruption. It would erase almost all known cognitive population level differences of class race etc. and it would potentiate an extreme increase in total human cognitive potential. I had not been aware that this result had been published, instead I was under the impression that an IQ uplift of ~3 IQ points with embryo selection was the current limit of intelligence enhancement available. Honest discussion about this approaching profound disruption in human civilization would seem highly appropriate and urgently required.

  236. Factorize says:

    Has the UKBB been factorized analyzed for intelligence (phenotypically, i.e., aside from genetically)? I was unable to determine this from the article. It will be exciting to watch this research unfold and witness the resulting changes in how intelligence is conceptualized.

Current Commenter
says:

Leave a Reply - Comments on articles more than two weeks old will be judged much more strictly on quality and tone


 Remember My InformationWhy?
 Email Replies to my Comment
Submitted comments become the property of The Unz Review and may be republished elsewhere at the sole discretion of the latter
Subscribe to This Comment Thread via RSS Subscribe to All James Thompson Comments via RSS