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
Adopt a Child, Drop an Illusion
Effects of adoption on intelligence: 42% heredity, 8% environment?
Email This Page to Someone

 Remember My Information



=>

Bookmark Toggle AllToCAdd to LibraryRemove from Library • B
Show CommentNext New CommentNext New ReplyRead More
ReplyAgree/Disagree/Etc. More... This Commenter This Thread Hide Thread Display All Comments
AgreeDisagreeThanksLOLTroll
These buttons register your public Agreement, Disagreement, Thanks, LOL, or Troll 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

I don’t do policy, but how about this one? In addition to all public policies aimed at getting rid of the achievement gaps between different groups, why not take an intensive approach? Continue with every program which is already under way, but add this one.

Get every child who is under-performing to live permanently from an early age with an adoptive family. This will give the children a deep immersion in all the things that good families can give their own children. We will carefully select adoptive parents who really want to adopt, many of which will have no children of their own, but very much want to bring up children.

This is an extreme experiment, far more profound than having just a few hours of enriching experience at a nursery. It will affect every waking moment, every spontaneous incident and remark, each instant of togetherness. Every shared experience can be used to teach, train and pass on the insights of wealth and status. How much will this boost IQ? To give this experiment every chance of showing its effects, we will wait till the children have grown up, and are 30 years of age. By now they should be well-established into their upward trajectories as a consequence of their enriched upbringing.

We do not actually have to do this experiment, many researchers have followed up adoptees. Here is the most recent study.

Genetic and environmental contributions to IQ in adoptive and biological families with 30-year-old offspring. Emily A. Willoughby, Matt McGue, William G. Iacono, James J. Lee. University of Minnesota Twin Cities, Department of Psychology, Minneapolis.

Intelligence 88 (2021) 101579
https://doi.org/10.1016/j.intell.2021.101579

The authors say:

While adoption studies have provided key insights into the influence of the familial environment on IQ scores of adolescents and children, few have followed adopted offspring long past the time spent living in the family home. To improve confidence about the extent to which shared environment exerts enduring effects on IQ, we estimated genetic and environmental effects on adulthood IQ in a unique sample of 486 biological and adoptive families. These families, tested previously on measures of IQ when offspring averaged age 15, were assessed a second time nearly two decades later (M offspring age = 32 years). We estimated the proportions of the variance in IQ attributable to environmentally mediated effects of parental IQs, sibling-specific shared environment, and gene-environment covariance to be 0.01(0.00-0.02], 0.04 [0.00-0.15], and 0.03 [0.00-0.07] respectively; these components jointly accounted for 8% of the IQ variance in adulthood. The heritability was estimated to be 0.42 [0.21-0.64]. Together, these findings provide further evidence for the pre-dominance of genetic influences on adult intelligence over any other systematic source of variation.

While adoption studies have provided key insights into the influence of the familial environment on IQ scores of adolescents and children, few have followed adopted offspring long past the time spent living in the family home.

So, bright parents create bright environments (add 1%), brothers and sisters in that family contribute to that environment (add 4%), and the interaction adds another 3%, thus 8% in all. However, if we look at the confidence limits (in brackets above) the low-ball figures could conceivably be nothing for the environment and 21% for genetics. The high-ball estimates would be 24% for environment, 64% for genetics. An even larger sample of adopted children would reduce those confidence limits.

At the Minnesota Center for Twin and Family Research (MCTFR), the Sibling Interaction and Behavior Study (SIBS) has followed a sample of adoptive and biological Minnesota families for nearly two decades. Initial IQ assessments were conducted when offspring were approximately 15 years of age, and this paper reports on new assessments taken at approximately 30 years of age. At age 15 children will still be under family influence, by 30 they will have established their own lives.

A recent development which transforms adoption research is to include entirely gene based polygenic risk scores. These are the new kids on the block, which provide the first approximation to a pure predictive genetic score. At the moment they are not able to capture all the effects which we know are due to heredity, as shown in twin research, but they are steadily increasing in power.

Our polygenic scores for educational attainment provide what is, at moment of writing, the largest R 2 estimates for any cognitive phenotype (0.113 for Total IQ; 0.154 for Verbal IQ). The particularly high predictive validity for PGS EA for verbal IQ is perhaps to be expected given that verbal IQ correlates more strongly than other IQ subscales with educational attainment, particularly for parents. These scores also enable a unique test for the so-called “placement effect,” wherein adoptees (typically twins reared apart) are thought by some skeptics to resemble their adoptive parents prior to placement, thus biasing biometrical estimates. By demonstrating a total lack of evidence (p =0.514) for a correlation between parents and adoptive offspring in polygenic scores, we provide support for the validity of at least some adoption studies in establishing causal inference.

That is to say, the idea that adopted kids are placed with adopting parents of apparently similar ability is not supported by these findings. (Critics said that the finding that identical twins reared apart correlated in intelligence was not due to their genetics, but to them having been placed in similarly bright families).

On the very important question as to whether adoption has a long-term impact on the ability of adopted children, the answer appears to be that no such effect can be found.

By examining parent-offspring resemblance in a sample of offspring that are among the oldest of any adoption study of IQ to date, we have effectively tested for the presence of parenting effects that would have persisted for more than a decade after the conclusion of the typical rearing period. No such persistence is found to occur in our unique sample.

This is pretty clear. It is unlikely that adoption has long-term effects on adopted children’s intelligence. This should give pause to all those proposing interventions less intensive than the full time parenting provided by adoption and expecting great results.

 
• Category: Science • Tags: Adoption, Heredity, IQ 
Hide 77 CommentsLeave a Comment
Commenters to Ignore...to FollowEndorsed Only
Trim Comments?
    []
  1. This should give pause to all those proposing interventions less intensive than the full time parenting provided by adoption and expecting great results.</blockquote

    A philosophical bre/ak/e for the woke debaters in this field.

  2. dearieme says:

    Critics said that the finding that identical twins reared apart correlated in intelligence was not due to their genetics, but to them having been placed in similarly bright families

    So the critics were conjecturing that the experienced social workers who organise adoptions were convinced hereditarians who believed that matching bright to bright was how they could best simulate nature’s ways.

    So unless the criticsy thought that the experience of such social workers was worthless, they were paying a cack-handed tribute to hereditarianism. Were they really so dim as not to recognise this? Or did they just declare, in the Marxist fashion, that the social workers were guilty of false consciousness?

    • Replies: @dearieme
    , @Curmudgeon
  3. dearieme says:

    these components jointly accounted for 8% of the IQ variance in adulthood. The heritability was estimated to be 0.42

    And the other 50%? It seems to me to be misleading to call it “environmental”. What should it be called: what’s a good adjective for a catch-all category for things not understood? “Stochastic” won’t do. I have it: call it “unknown”.

    The massive contribution of genetic inheritance is established beyond doubt. It doesn’t really matter if the “unknown” is 50%, or perhaps only somewhere above about 12% (100% – 64% – 7% – 15% – 2% ), or maybe a bit below 79% (100% – 21%) . It’s surely got to be the target for almost all further research.

  4. dearieme says:
    @dearieme

    criticsy was a typo but carries a pleasant implication of critics + clerisy.

  5. David says:

    Prospero experienced this:

    A devil, a born devil, on whose nature
    Nurture can never stick; on whom my pains,
    Humanely taken, all, all lost, quite lost;
    And as with age his body uglier grows,
    So his mind cankers.

    • Thanks: dearieme
  6. BCB232 says:

    My eye-brain combo wants to draw the slope on the left graph steeper but who am I to argue with statistical software.

    The slope of the regression line on the left graph isn’t one to one. Is that the effect of regression to the mean?

    Or more low-IQ outlier situations like profound mental retardation from mutations like trisomy?

    • Replies: @Emily W
  7. Ray says:

    The most recent adoption in which I was involved was around 2013 or so .
    It was then accepted in English law that adoptive parents must be given full information about their child’s background – including information about the birth parents and family -when there was any chance of it affecting the child and or the adoptive family. This included illnesses where there was or is or could be a familial pattern – diabetes, heart disease, Alzheimer’s , hypo-tension etc; psychiatric factors- bipolar, schizophrenia, ADHD , – and parental intelligence. ( Usually the birth-parents are pretty thick – quite a few of them were nasty in all senses of the word. )
    It was accepted – at least tacitly- that the intelligence of the birth parents was quite predictive of the child’s development. And we tried to convey to future adopters that no matter what they did the teenage years of their new child were bound to be -well – interesting . Or whatever euphemism came to mind. Usually the prospective adopters were kind, intelligent and loving . We tried to persuade them that their own characteristics are absolutely NOT guaranteed to be acquired by their child. But on the other hand , most people do not acquire a family pet because they imagine that it will go to university or even get a prize at Cruft’s.

    • Replies: @dearieme
    , @Colin Wright
  8. @dearieme

    As an aside, a late uncle was a social worker. By the early to mid 60s, some of his classmates had moved into university positions. In the early 80s, I sat on a committee guiding an employee assistance plan. One of the committee turned out to be one of his old classmates, who, according to my uncle, was one of the “new wave” types to whom a spade wasn’t a spade, it was an agricultural implement. Long on hypothesis, short on reality.

  9. dearieme says:
    @Ray

    information about the birth parents

    Insofar as the father is known.

  10. Wency says:

    So, bright parents create bright environments (add 1%), brothers and sisters in that family contribute to that environment (add 4%), and the interaction adds another 3%, thus 8% in all.

    I’m just curious, is this generalizable? Is there an argument for family size here? It sounds like it’s saying that creating a larger number of high IQ children (up to a point, recognizing deleterious effects of parental age also) will reinforce the adult IQ of the others, and this effect is far more powerful than anything the parents themselves can do directly (while still tiny compared to genetic effects).

    • Replies: @James Thompson
    , @Emily W
  11. @Wency

    Considering the confidence levels, I wouldn’t bet on it.

    • Thanks: Wency
  12. dearieme says:

    If nurture conquers all, no bright parents would ever have dim children, nor dim parents bright. You need the genetic lottery to explain such ill-matched generations.

    With this simple argument I once persuaded a journalist to give up her extreme nurturist views. “Why” she asked “have I never heard this argument before?”

    “Because you drink with duds” said I, cheerfully.

  13. @Ray

    ‘…It was accepted – at least tacitly- that the intelligence of the birth parents was quite predictive of the child’s development. And we tried to convey to future adopters that no matter what they did the teenage years of their new child were bound to be -well – interesting . Or whatever euphemism came to mind. Usually the prospective adopters were kind, intelligent and loving . We tried to persuade them that their own characteristics are absolutely NOT guaranteed to be acquired by their child. But on the other hand , most people do not acquire a family pet because they imagine that it will go to university or even get a prize at Cruft’s.’

    I see some of that around here. The area is almost perfectly non-black (0.3% black), but there is the occasional black child, apparently adopted out of an excess of Christian zeal.

    It never looks like it is working out.

    • Replies: @YetAnotherAnon
  14. And gee, there couldn’t be any flaws in these surveys or anything, now could there. Science is perfect.

  15. Travis says:

    Imagine if Steve Jobs was adopted by a hillbilly family in West Virginia ?

    Really poor and dysfunctional families rarely get to adopt kids, but it would be interesting to see the outcome if mulatto children were adopted by Black parents with 90 IQs compared the the mulatto children adopted by Whites parents with 100 IQs. Probably not many parents with two digit IQs adopt many children.

  16. dearieme says:

    Does anyone study the effect on IQ (and other traits) of fostering?

  17. anonymous[184] • Disclaimer says:

    “Imagine if Steve Jobs was adopted by a hillbilly family in West Virginia ?”

    You’d be able to watch porn on your favorite I-rifle during deer season.

    • LOL: dearieme, AceDeuce
    • Replies: @Stargazer
  18. Anonymous[333] • Disclaimer says:

    …high predictive validity for PGS EA for verbal IQ is perhaps to be expected given that verbal IQ correlates more strongly than other IQ subscales with educational attainment…

    Verbal IQ subscale needs to be completely reworked. It measures vocabulary more than intelligence which is exactly the wrong way to measure real IQ. Every other component is, rightfully, eliminating the influence of knowledge differences as much as possible but verbal IQ is actually going in the opposite direction for some inexplicable reason.

  19. Stargazer says:
    @anonymous

    anon(184): Now that was funny!

  20. In fact, Vocabulary is a very good test of intelligence. It is not about hearing words. It is about learning in what contexts and with what purposes they are to be used.
    I did quite a few post on this a few years ago, which you might find by puting Vocabulary in my search bar.

    Here is one to start with, which makes my main points

    https://www.unz.com/jthompson/vocabulary-humanitys-greatest

    • Replies: @Sollipsist
    , @Anonymous
  21. Ray says:

    Re vocabulary and IQ .

    I can see the possible link between vocabulary size and intelligence , subject to the usual caveats .

    I wonder though if for example having an average vocabulary in two or three languages is the equivalent of a large vocabulary in only one language ?
    Just asking for a friend

    • Replies: @James Thompson
  22. @James Thompson

    In which case, the characteristic decontexualization that the Internet provides is possibly the worst thing that we could have done to support intelligence development.

  23. Anonymous[130] • Disclaimer says:
    @James Thompson

    Sorry, but we’ll have to disagree on this one. A proper IQ test should eliminate knowledge (or knowledge differences, to be exact) as much as possible when measuring intelligence. That’s how it works with other components. A “math” question in a normal IQ test would always stick to basic math so it wouldn’t matter if the test-taker had a high-school or university diploma.

    Verbal IQ subscale is the only one going in the opposite direction. It literally favours a CRT dullard over a Physics genius who might not know, or care about the finer points of “intersectional mansplaining”.

    • Replies: @Rosie
    , @James Thompson
  24. @Ray

    Can’t remember at the moment, but think that competence in several languages is intelligence related. Don’t think that average vocabulary in several languages would be anything like high vocabulary ability in one language.

    • Replies: @Ray
  25. res says:

    That is to say, the idea that adopted kids are placed with adopting parents of apparently similar ability is not supported by these findings. (Critics said that the finding that identical twins reared apart correlated in intelligence was not due to their genetics, but to them having been placed in similarly bright families).

    Thanks for highlighting that subtlety. An interesting application of the PGS.

    Just to make sure I understand correctly, in Figure 2 at the top the two panels contain the same children just with the different sets of parents, right? I found the wording about biological and adoptive families unclear as to whether the children were all the same. Probably clarified somewhere in the text. Or should it be obvious?

    It would probably have too much clutter to be very useful, but would it be possible to create versions of Figure 2 which connect the intake and follow-up scores (e.g. a line between them)? Perhaps separate out the cases where the scores differ greatly to avoid long lines running through all of the other data?

    Did they look at age of adoption as a factor?

    Would it be of any use to look at height in a similar fashion for the same group and compare the results? I think it might be interesting to compare residuals from regression and prediction and see if height and IQ tend to vary in the same direction for those.

    P.S. Is full text available anywhere?

    P.P.S. James Lee has been busy. This and EA4 around the same time?!

    • Replies: @Emily W
  26. Rosie says:
    @Anonymous

    Verbal IQ subscale is the only one going in the opposite direction. It literally favours a CRT dullard over a Physics genius

    In theory, maybe, but in practice?

    Unless there is new evidence to the contrary, wordsum correlates better with intelligence than educational attainment. This rings true to me. At least in previous generations, it was not uncommon for very intelligent working-class people to have relatively low educational attainment. They read the newspaper and did crossword puzzles (associative reasoning) while talking and sharing interesting things with their granddaughters.

    https://www.researchgate.net/figure/Distribution-of-raw-scores-on-the-Wordsum-test_fig1_271074326

    • Replies: @Anonymous
  27. Rob says:

    Did the adptees get an IQ boost a whole? If adoptees were + 15 IQ (likely smaller) over their expected iq if the hadn’t beem adopted, then thst would not show up in the correlation, right?

    Also, it’s curious thst the correlations’ confidence interval cut off a zero. Correlation (ρ) can go negative. I would think the confidence interval for a ρ of 0.01 would have a confidence interval that went negative. Is my intuition off? Did they carefully chose a CI percent that would stay positive.

    In my experience, it seemed as if smarter people often have trouble explaining concepts to considerably dumber people, so I could see that 120 IQ parents might have trouble explaining things to 85 IQ kids. Is my intuition off, yet again. Maybe it is. I seem to recall that white teachers do better teaching black kids than black teachers. That is likely to be a 15 point gap, if not more.

    • Replies: @RogerL
    , @Emily W
    , @Philip Owen
  28. RogerL says:
    @Rob

    Rob,
    I have a some observations about why brighter people often have trouble explaining things to less bright people.

    A lot of people have memorized a lot of jargon, and the simpler interdependencies between the concepts. However they don’t clearly understand the essence of those concepts and processes, so they have trouble translating it into everyday language.

    The IQ concept doesn’t distinguish between numerical and verbal aptitudes. Translating concepts to a different dialect is a verbal reasoning skill, and the skills of a lot of bright people are centered on numerical reasoning, not verbal reasoning.

    There is an unwritten guild law that says thou shall speak in jargon and thou shall be ejected into the wilderness if you speak in everyday language. “Experts” who speak in everyday language are often attacked by their peers, so most don’t ever try to do this. Any aptitude a person has, which isn’t ever exercised, atrophies, and then they can’t do it even if their life depended on it.

    ~
    That said, good teachers work hard at finding ways to reach all students, so likely they have more practice translating concepts to different dialects. As a result of this practice, some teachers get pretty good at it.

  29. Anonymous[130] • Disclaimer says:
    @Rosie

    In both theory and practice you can teach a person to score much higher on the verbal IQ component by giving them a number of select books and a thick thesaurus.

    That’s a huge red flag. The person didn’t increase his intelligence (raw brain power – not knowledge) but his official IQ is now higher. As a matter of fact, he can continue increasing his verbal IQ score all the way into the old age as his actual physical brain starts deteriorating and other IQ subscales decrease. That’s another red flag.

    The fact that there is some correlation in many cases is irrelevant. You’re not supposed to measure knowledge in a proper IQ test – ever – and vocabulary is knowledge that varies wildly across time and space. James Thompson already published a number of articles showing IQ correlation with career advancement or even health outcomes but nobody is suggesting that salaries or health records should be incorporated into IQ tests. There’s a reason for that.

    • Replies: @Colin Wright
    , @RogerL
  30. @Anonymous

    As a general rule, I would agree with you on the principle that intelligence test questions should require only the knowledge that everyone can pick up by being a member of society, and avoid the more specialised knowledge that might be imparted by a particular school curriculum. However, the empirical finding is that vocabulary seems to fall in the first, generalist category, and not the specialist one. No valid, general population vocabulary test would major on highly specialist fields, so that would cut out intersectional mansplaining. These general vocabulary tests show a close correlation with other intelligence tests, and with real life achievements.

    • Replies: @Anonymous
  31. RogerL says:

    This is one of those topics where I followed the threads to a surprise conclusion (at least it was surprising to me).

    It helps the gut level belief in this to see an adoption in action.

    My aunt and uncle lived on the east side of San Francisco bay. He designed instrumentation for managing industrial processes. She was the head of a high school English department. Pretty typical suburban, upper middle class.

    They had 2 daughters, my uncle wanted a son, and didn’t want to end up with many daughters in an effort to have a son. So they adopted a son. The son turned out to be a classic good old boy – large, slightly overweight, friendly, easy going and outgoing. Out of high school he ran the barbecue concession at a local park. Eventually he became a refrigeration technician. So he wasn’t stupid, but he wasn’t up to doing the engineering to design what he worked with. More importantly, he liked interacting with people and fixing things, and he hated paperwork. The more my cousin resisted joining the professional-worker social-class, the more they all treated him like a deplorable. Then he moved to northern Idaho where he fits in better.

    There must be hundreds of millions of kids with screwed up lives because the adults raising them believed that kids are clay and they could make the kids into anything they wanted, while at the same time the kids were feeling in their bones what they were, and weren’t. This includes ambitious working-class families that “give their kids every advantage” in an effort to force them into the upper middle class, where they don’t fit, and don’t want to be.

    A few times I lived with that aunt and uncle for a while, which reinforced my west coast cultural characteristics. There is a huge cultural gap between the Bay Area and New York City. These are fundamentally very different cultures. How much of this cultural difference is absorbed during childhood, and how much of it is genetic? Did the people with a genetic inclination to be easy going and freewheeling move to the west coast and reinforce and magnify the difference between the west coast and east coast of North America?

    In any case, its obvious that I had engineering and management genes that my cousin didn’t, and more than simple aptitude, I had an interest in doing those activities, and he didn’t.

    ~
    A related issue, which I see very little discussion on, is the effects of mixed marriages on the children. I don’t see any discussion of the effects of this on society.

    Since a lot of behavior is genetically based, children from mixed marriages don’t fit into any cohesive culture because some of their behavior is problematic in any one culture. Children from mixed marriages are permanently handicapped in their ability to live a successful life. Some succeed in spite of this handicap. How much more successful would they have been without that handicap?

    Maybe there are gumbo cultures somewhere, which have a bit of everything in them, but I doubt if they are cohesive, successful, or enduring.

    There are creole cultures. However, once they pass thru the mixing and evolution processes and have a cohesive culture, the parts of the parent cultures, which were dropped, don’t fit in any more with the creole culture. So once again there is a clear concept of what is “our” culture, and what is culture from other cultures, which is excluded from their culture. It seems highly likely that the remaining cultural parts reinforce the other parts in constructive ways, and the parts from the parent cultures, which would cause cognitive dissonance within the creole culture, were dropped.

    So a marriage between a creole person, and a person from one of the originating cultures, would be still be a mixed marriage, and their children would have problems fitting into either the parent culture or the creole culture because some of their behavior genes would be at odds with that culture.

    A very common example of a fundamental and chronic cultural behavior conflict is that in the professional-worker social-class loud raucous laughter is never okay in any setting, public or private, and many other cultures think life is flat and boring when loud raucous laughter is not allowed. This particular issue is a root in many severe cultural conflicts.

    People, who promote mixed marriages, are right up there with the Nazi doctors who did destructive experiments on humans.

    How much of the conflict going on in the west today is caused by people who are too genetically mixed to fit in anywhere, are deeply unhappy, and are failing around?

    This genetic mismatch is complicated by the fact that many people use defense mechanisms to avoid and bury issues they don’t want to deal with, and then they project outwards their unresolved distress.
    https://en.wikipedia.org/wiki/Defence_mechanism

    How many of the vicious woke are projecting outwards their own distress about not fitting in, and in true modern fashion are blaming others for their sense of distress?

  32. Emily W says:
    @BCB232

    The slope of the regression line on the left graph isn’t one to one. Is that the effect of regression to the mean?

    Two reasons: First, for intake IQ of the parents, we only had mom’s or dad’s IQ for each family, but not both. For follow-up 3, we had ICAR-16 data for both parents in the majority of cases. Second, there was some minor attrition with a handful of families present in intake but not in follow-up. Attrition wasn’t high enough to affect robustness of the comparison (see supplement for details).

    Or more low-IQ outlier situations like profound mental retardation from mutations like trisomy?

    We had no profoundly low-IQ individuals in this data set. In order to ensure full understanding of the written materials, participants at intake (long before my time) were screened for IQ > 70, and only one participant was dropped from intake data as a consequence.

    • Replies: @res
    , @BCB232
  33. Emily W says:
    @Wency

    I would err on the side of caution and say that these conclusions have limited generalizability. None of the families we studied came from extremely impoverished backgrounds, and all grew up in a relatively homogenous culture (Minnesota). As for family size, the adoptive and biological families we included were matched to each other for number of offspring, typically two in each family. They were also at least roughly matched for children’s ages.

    • Thanks: Wency
  34. Emily W says:
    @res

    Just to make sure I understand correctly, in Figure 2 at the top the two panels contain the same children just with the different sets of parents, right?

    The two panels represent mostly different parents and offspring, but there is some overlap: Some families were mixed, in that they had both adoptive and biological children. Figure 2 is meant mostly for illustrative purposes—the biometric decomposition took non-independence into consideration.

    Did they look at age of adoption as a factor?

    All adoptees in this cohort were required to have been placed for adoption before reaching two years of age (M = 4.7 months, SD = 3.4 months). We did not explicitly look at the age of adoption within this range as a predictive factor, though I probably have that data lying around somewhere…

    P.S. Is full text available anywhere?

    You can find the full text and supplement on James Lee’s website (first entry under 2021): https://jamesjlee.altervista.org/

    • Replies: @res
  35. Emily W says:
    @Rob

    Did the adptees get an IQ boost a whole? If adoptees were + 15 IQ (likely smaller) over their expected iq if the hadn’t beem adopted, then thst would not show up in the correlation, right?

    Although this wasn’t the focus of the paper, adoption did appear to raise the children’s IQ by an average of 2.83 points.

    Also, it’s curious thst the correlations’ confidence interval cut off a zero. Correlation (ρ) can go negative. I would think the confidence interval for a ρ of 0.01 would have a confidence interval that went negative. Is my intuition off? Did they carefully chose a CI percent that would stay positive.

    The familial correlations’ CIs weren’t cut off at zero. You can find the 95% CI for each familial correlation in the supplement (Table S11): https://jamesjlee.altervista.org/wp-content/uploads/2021/08/SI_IQ_adoption_rv2_final.pdf Indeed, some of them dipped well below zero, especially for the adoptive relationships.

    The CIs for the variance estimates (main text table 3) were cut off at 0, because variance can’t be negative. (Incidentally, negative C variance can result from biometric models when there’s evidence for non-additive genetic variance, but that wasn’t the case here.)

    • Thanks: Rob
    • Replies: @res
  36. res says:
    @Emily W

    The slope of the regression line on the left graph isn’t one to one. Is that the effect of regression to the mean?

    Two reasons: First, for intake IQ of the parents, we only had mom’s or dad’s IQ for each family, but not both. For follow-up 3, we had ICAR-16 data for both parents in the majority of cases. Second, there was some minor attrition with a handful of families present in intake but not in follow-up. Attrition wasn’t high enough to affect robustness of the comparison (see supplement for details).

    Thanks for the explanation. I think this text from just below Table 1 captures the relevant information.

    Valid measures of IQ are available from this intake sample for 461 mothers, 46 fathers, 690 adoptive and 538 biological offspring. No information on the biological parents was available for all adopted offspring.

    It is disappointing (but understandable) not to have any biological parent IQs for the adopted offspring.

    I would have expected having the IQ for both parents at follow-up to give a stronger correlation (opposite of what was seen). Similarly for the presumably higher IQ heritability in the older children. Any thoughts on that? Did you try any mixing and matching of parent/child intake/follow-up IQ scores to see how the correlations differed? The full matrix seems like it would be of interest. That sounds a bit like Tables S8-S10, but is different. More like Table S11 which if I understand correctly gives the intake and follow-up correlations as Total IQ and ICAR-16 (but I would add things like Mom Total IQ and bio ICAR-16). Given the Figure 2 r=0.42 and Table S11 Total IQ correlations of 0.36/0.46 for Mom/Dad and bio (and remembering that mothers were heavily overrepresented for intake IQs) I assume Table S11 is only including those who followed up and those numbers (0.42 vs. something like a weighted 0.37) are showing the attrition effect you mentioned?

    This from page 8 was also interesting.

    The ethnic composition of offspring in this sample is additionally unique; while 95 percent of parents and biological offspring are of non-Hispanic white European ancestry, 21 percent of the adoptive offspring are white, 66 percent are Asian, and 13 percent are of other ethnicities

    How comfortable are you with using the EA3 PGS for Asians given that it was developed with Europeans? I see significant effort to look at that in both the paper and the SM. Table S14 is the best summary of that I see. There we see the R^2 for PGS_EA reduced about 60% and 1/3 for Total IQ and ICAR-16 for Asians relative to whites. But the most interesting result is the Years of education R^2 for Asians being twice that of whites?! Any idea why that would be?

    P.S. This turned into a bit of a question bomb. Sorry. I appreciate any elaborations you care to make.

    • Replies: @Emily W
  37. res says:
    @Emily W

    Thank you! Much more interesting material at that link as well (e.g. the book chapter from you and James Lee looks like a nice primer on research methods in this area, the Not by g alone paper looking at college educated people with IQ < 90 is relevant to some NLSY conversations we have had here).

    The Figure 2 clarification is very helpful. I had badly misinterpreted it.

    That is a pretty narrow range for age of adoption. Seems likely not to be an issue given that?

  38. res says:
    @Emily W

    Incidentally, negative C variance can result from biometric models when there’s evidence for non-additive genetic variance, but that wasn’t the case here.

    That is an extremely interesting point. Could you elaborate on that a bit? Will the presence of non-additive genetic variance always (or usually) lower C below its true value? With that variance ending up in E for an ACE model? Or would it appear in G-E covariance?

    Are any traits well enough understood (e.g. relatively known magnitude of non-additive genetic variance) to take a quantitative look at this effect? Does this happen for “reasonable” magnitudes of non-additive vs. additive genetic variance?

    • Replies: @Emily W
    , @Rob
  39. @Anonymous

    ‘In both theory and practice you can teach a person to score much higher on the verbal IQ component by giving them a number of select books and a thick thesaurus.’

    I question the truth of that.

    • Replies: @Anonymous
  40. Anonymous[684] • Disclaimer says:
    @James Thompson

    “Intersectional mansplaining” was just a crude example. The point was that CRT dullards love their “big” words to the point of feverishly inventing new ones on a daily basis. Their lectures and literature will be much denser, in more ways than one, when compared to Math or Physics studies.

    My main point are:

    – A good IQ test should always put every test-taker on a level playing field in terms of knowledge. Vocabulary knowledge in particular varies wildly across age groups, educational levels, chosen subjects, schools, no schools and above all across different temperaments.

    – We also have good correlation between school grades and intelligence but we don’t inject them in IQ test scores for a good reason.

    – Verbal IQ is the only subscale that can be boosted through learning – even in old age as the physical brain and other subscale measurements start to decline. That’s just laughable. We clearly have a problem here. One of these things is not like the others.

    • Replies: @James Thompson
    , @res
  41. Anonymous[684] • Disclaimer says:
    @Colin Wright

    Why? Verbal IQ subscale relies a lot on vocabulary (size). Vocabulary is a type of knowledge that can be learned. As a matter of fact, most of us will expand our word libraries until death or senility kicks in.

  42. RogerL says:
    @Anonymous

    You are correct that a poorly implemented vocabulary test would measure education, not intelligence. In addition, a person taking an intelligence test, which isn’t in their first language, would be handicapped, and they wouldn’t score as high in verbal intelligence as they would on a test written in their first language.

    All that said, it takes verbal intelligence to understand the significance of the subtle nuances between synonyms. In the intro for this page, it gives an example of a set of synonyms: begin, start, commence, and initiate
    https://en.wikipedia.org/wiki/Synonym

    While they are synonyms, they also have different nuances. The better a person understands those nuances, the stronger their verbal intelligence will be. While the word “commence” isn’t commonly used, it seems to me that the other 3 words are widely used, and most people would have an opportunity to pick up on the distinctions between them, if they have the verbal intelligence to do that.

    If a person doesn’t have the verbal intelligence to see, understand, and articulate the nuances in how these words are different, then it seems unlikely that they could study up and increase the score on their IQ test.

    Probably, as they get older, some people gain a better understanding of these nuances. A compensation would be to compare scores for people of the same age group.

    There are all kinds of pitfalls with implementing intelligence tests, and possible mitigating factors for individual results. For instance, a person who doesn’t read, and lives in isolation with very little verbal interaction with others, will have a stunted verbal intelligence score.

    The key point is that well designed tests for verbal intelligence provide some useful general indications for most people. However if in a specific case, exactness is needed in assessing a person’s intelligence, then more in-depth testing should be done.

    It seems like there are 2 different things. One is the ideal for intelligence tests that the creators of the tests strive for. The other is recognizing when intelligence tests are good enough to provide useful results, and also recognizing their limitations and when the results won’t be useful.

    • Thanks: res
  43. @Anonymous

    I had better look at your references showing that verbal IQ can be boosted through learning. Can you send me a few key ones?

    • Replies: @Anonymous
  44. Anonymous[684] • Disclaimer says:
    @James Thompson

    Sure, let’s see some references:

    1) In this article you’re stating:

    verbal IQ correlates more strongly than other IQ subscales with educational attainment

    That’s a bit of an understatement, in my opinion, but let’s just say that verbal IQ score can be improved through “educational attainment” – a.k.a. learning. Red flag. You’re not supposed to learn your way into higher IQ score – especially if your physical brain cells start deteriorating and every other IQ measurement declines.

    2) Your reply #20 links to this article where you state:

    As a rough guide, teenagers have about 12,000 words and college students 17,000. Older adults have 17,000 to 21,000 words, and a minority have many more.

    Those are huge differences – and they have nothing to do with intelligence and everything to do with learning. It’s also a “rough” guide which completely ignores individual choices of schools, hobbies or chosen educational pathways and goals. I’ve already noted how some of the dumbest “social sciences” fill their dogma programme with unnecessary exotic words for the sake of appearing smart. I’m willing to bet big money – without reading – that “White Fragility” contains many, many more “big” words than any Physics book of similar size.

    3) I distinctively remember one of your own articles referencing a study – tracking lifetime IQ measurements – showing verbal IQ trending upwards – up to a point – while every other IQ measurement and the average efficiency of physical brain cells declined. Again, one of these things is not like the others and that’s just laughably wrong.

    Unfortunately I don’t have a link but I’m sure you can find it. Bottom line is that data in this instance doesn’t matter. We can use reason. Please explain to me why an average 60-year-old can’t expand his vocabulary by, say 50% if he’s dedicated, and significantly increase verbal IQ score. This could easily overtake any actual brain cell deterioration so please explain to me what we’re measuring with verbal IQ? We can literally have a case where a deteriorating, physically sub-optimal brain can improve its verbal IQ score. Why is that allowed?

    • Replies: @res
  45. BCB232 says:
    @Emily W

    Emily W. thanks for the response. It’s pretty cool and a privilege to be able to interact with one of the study authors.

    • Agree: res
  46. @Colin Wright

    The blues musician John Mayall and his then-wife Pamela adopted a Nigerian infant (Benedict) in the 1960s – I don’t know if that was John’s attempted payback for the blues music that made his career (as well as that of Clapton/Peter Green/Mick Taylor etc). Then they were divorced in the early 1970s.

    https://www.alamy.com/stock-photo-john-mayall-of-john-mayall-the-bluesbreakers-with-his-wife-pamela-84178451.html

    In the 1980s I used to go to his son Gary’s “Gaz’s Rockin’ Blues” dance nights in London. It’s still going!

    https://www.gazrockin.com/gass-rockin-blues

    I wonder what happened to Benedict Mayall?

  47. res says:
    @Anonymous

    The point was that CRT dullards love their “big” words to the point of feverishly inventing new ones on a daily basis. Their lectures and literature will be much denser, in more ways than one, when compared to Math or Physics studies.

    Hard to argue with that ; )

    A good IQ test should always put every test-taker on a level playing field in terms of knowledge.

    I think you need a qualifier in that. Say “as much as possible”? You do realize as stated your goal is unattainable, right?

    Verbal IQ is the only subscale that can be boosted through learning

    That seems obviously wrong. Do you really think practicing Raven’s Progressive Matrices style questions would not help someone who was previously unexposed to that type of thing? At the extreme end, test prep which included the exact questions to be used later would presumably improve results on most tests.

    The question I see is how much do learning issues matter in practice for IQ testing? RogerL’s comment 43 provides a good take on that.

    Some dimensions to consider here.

    1. Variability of performance with learning. How much of a difference does the full range (say raised by wolves to reading voraciously since age 3) of learning variability make for IQ? How much of a difference does the typical range (say 5% to 95% of reading/learning effort in a population) of learning variability make for IQ? How much effort is required to move measured IQ? I would expect the improvement curve would be steep for much of going from raised by wolves to mandatory K-12 education, but much shallower after that.

    2. Persistence of learning improvements on measured IQ. I would expect specific training (e.g. tell people all of the words in Wordsum) to fall off quickly while general learning would be more durable (but also less dramatic in effect).

    • Replies: @Anonymous
  48. res says:
    @Anonymous

    verbal IQ correlates more strongly than other IQ subscales with educational attainment

    That’s a bit of an understatement, in my opinion

    If we want to make this discussion quantitative Table S8 contains:

    Table S8: Correlation matrix for all valid individual scores of IQ and IQ subscales, years of education, highest degree obtained, EduYears polygenic score, and ICAR-16 score. Only measures shared across parents and offspring are included; i.e., all lQ scores and subscales are those taken at intake, while educational attainment, SES and ICAR-16 measures were evaluated at follow-up 3.

    Correlations with Years of education were:

    1. Highest degree .81
    3. SES .41
    4. PGSEA .25
    5. ICAR-16 .30
    6. Verbal IQ .38
    7. Performance IQ .18
    8. Total IQ .35
    9. Information .37
    10. Vocabulary .33
    11. Picture completion .11
    12. Block design .16

    Those are huge differences – and they have nothing to do with intelligence and everything to do with learning.

    That false dichotomy is a gross overstatement. I’m curious, do you think intelligence affects the ability/efficiency of learning?

    I’m willing to bet big money – without reading – that “White Fragility” contains many, many more “big” words than any Physics book of similar size.

    I hope you haven’t been foolish enough to actually place that bet with anyone (especially given the “many, many more” bit). But if you’d like to do it now… I took roughly 3000 word excerpts from Chapter 1 of White Fragility and Fundamentals of Physics (Resnick, et al., had to edit significantly to remove the complexity of equations, etc. ; ) and ran them through https://readabilityformulas.com/free-readability-formula-tests.php

    The results were surprisingly similar overall. I thought 3+ syllable words was the best benchmark of what you asserted.

    White Fragility

    Total # of words: 3284
    Total # of unique words: 879 (27% of total text)
    Total # of repeat words: 2405 (73% of total text)
    Percent of 3+ syllables in text: 17%
    Total # of words with 3+ syllables: 566
    Grade Level: 12
    Reading Level: difficult to read.

    Fundamentals of Physics

    Total # of words: 3400
    Total # of unique words: 1036 (30% of total text)
    Total # of repeat words: 2364 (70% of total text)
    Percent of 3+ syllables in text: 16%
    Total # of words with 3+ syllables: 544
    Grade Level: 10
    Reading Level: fairly difficult to read.

    But if you look at the actual 3+ syllable words used (after the MORE) I think it is clear one is much more complex than the other. Though White Fragility is higher on grade level. Worth noting that the excerpts are not perfect (easily seen from the actual words below, with the physics version being a bit worse because of much non-word matter).

    tracking lifetime IQ measurements – showing verbal IQ trending upwards – up to a point – while every other IQ measurement and the average efficiency of physical brain cells declined. Again, one of these things is not like the others and that’s just laughably wrong.

    The 389 page WJIV Technical Manual is an excellent reference for age curves of different subtests.
    https://www.wjscore.com/Files/WJIVTechnicalManual.PDF
    More at this comment.
    https://www.unz.com/jthompson/no-skin-in-the-game/?showcomments#comment-4100684

    See Figures 5-3 and 5-4 on page 137. You are right about one thing being not like the others.
    The Comprehension-Knowledge (Gc) factor cluster and Vocabulary (VL/LD) curves both show the gradual increase from about age 20 to 65 you mention. The others show varying levels of flat and declining after age 20. Cognitive Processing Speed (Gs) and Perceptual Speed (P) are notable outliers in the other direction (declining more quickly than the others).

    Bottom line is that data in this instance doesn’t matter. We can use reason.

    Yes, we have often seen how well that works in the past. /sarc

    [MORE]

    White Fragility

    3+ syllable words: (show all ‘hard’ words)
    RACISM | OURSELVES | American | United | reference | experience. | experience | universal | experience. | particularly | experience | society | profoundly | society | separate | unequal | however | certainly | attention | somebody’s | critical | component | ability | discomfort | racially | racially | fragility | stamina | OPINIONS | UNINFORMED | opinion | racism. | possible | United | significant | history | colonization—without | developing | opinions | racism. | opinions | racism | relations | profoundly | consider | devoted | intentional | ongoing | opinions | necessarily | uninformed | ignorant. | opinions | racism | ignorant | information | understanding | arguably | enduring | dynamic | several | example | qualified | organization | understanding | whatsoever | perspectives | experiences | relationships | virtually | ability | critically | graduate | discussing | racism. | graduate | discussing | racism. | education | discussing | racism. | considered | progressive | “diversity” | faculty | overcome | resistance | majority | colleagues | diversity | various | guarantee | racism. | openly | honestly | fragility | defensiveness | argumentation | certitude | natural | responses | attaining | productively | powerfully | hierarchy | ideologies | individualism | meritocracy | repetitive | media | representations | segregation | neighborhoods | depictions | whiteness | truncated | history | openly | solidarity. | interrupting | racism | ongoing | lifelong | conditioning | frameworks | simplistic | definition | racism—as | intentional | discrimination | committed | immoral | individuals | —engenders | confidence | evidence | implausible. | example | everyone | oneanother | discussion | engagement. | unconvincing | invalidate | experiences. | understand | socialization | UNDERSTAND | SOCIALIZATION | racism | responses | predictable | reciting | significant | ourselves | objective | unique. | understand | fragility | understand | understand | socialization. | perceptions | experiences | particular | cultural | universal | objective | society. | exploring | cultural | frameworks | particularly | challenging | precisely | ideologies | individualism | objectivity. | individualism | objectivity | possible | ideologies | difficult | collective | experience. | individualism | communicates | individual | memberships | irrelevant | opportunities. | individualism | intrinsic | barriers | individual | consequence | individual | character. | according | ideology | individualism | irrelevant. | occupy | positions | profoundly | natural | voluntary | opportunity | equally | distributed | opportunities | benefit | mediocre | exceptional. | thoughGates’s | ideology | individualism | consider | regardless | protestations | everyone | dominant | different | experience | different | different | different | disability | different | heterosexual | naturally | difference | experience. | myriad | variety | mediums. | continues | nonverbal | comparing | ourselves | socialized | collectively. | different | experience | another. | opposite | example | disability | understanding | collectively | society | unavoidable | television | magazines | religion | literature | traditions | history | dimensions | identities. | understanding | ourselves | necessarily | comparisons | deserving | undeserving. | understand | understanding | society’s | emphasis | individuality | reflecting | memberships. | understand | relations | conditioning | memberships | addition | challenging | ourselves | individuals | identity | objectivity. | membership | relevant | universal | perspective | perspective | particular | ideologies | disrupted. | reflecting | particularly | challenging | unfortunately | denying | examine | important | remember | consider | socialization | difference | verbally | resistance | cardinal | individualism—I | generalizing. | proceeding | anything | different | neighborhood | experience | different—that | example | recently | organization | African | American. | emphasized | importance | humility | exempting | ourselves | unavoidable | dynamics | racism. | formed—ostensibly | questions—but | typically | reiterate | opinions | Italian | American | Italians | considered | discriminated | experience | racism | overwhelmingly | coworkers | examination | whiteness | Italians | discriminated | toocommon | example | individualism. | engagement | consider | Italian | Americans | assimilation | experiences | illustrate | different | similar | exception | precisely | separate | sociologist | comfortable | generalizing | predictable | measurable | understand | generalizations | defensiveness | generalizing | ideology | individualism | exceptions | recognized | precisely | recurring | predictable. | understand | racism | behavior | individuals. | specific | adjustments | necessary | situation | evidence | entirely. | example | poverty | Ashkenazi | European | military | family. | Canada | Hawaii | Germany | family. | situations | racism | society | exemption | examination | experience. | uniqueness | critical | society | individualism | individual | narrative | collective | SIMPLISTIC | UNDERSTANDING | RACISM | definition | post–civil | intentionally | immoral. | therefore | offensive | questioning | character. | generalize | according | definition | offensive | definition | racism | racism | definition | racism | immoral. | argument | discomfort | unsettle | comfortable | relations | comfortable. | discomfort. | messenger | disregard | unsettle | understanding | dynamics. | unexamined | assumptions | possible | dynamics | consider | possibility. | different | interrupt | fragility | capacity | discomfort | discomfort | racially | discomfort | humility. | understand | socialization | constantly | inability | acknowledge | inevitably | resistance | defensiveness | fragility. | stamina | fragility | identities—and | identity | particular. | struggling | whitedon’t | information | understanding | arguably | enduring | dynamic | several | example | qualified | organization | understanding | whatsoever | perspectives | experiences | relationships | virtually | ability | critically | graduate | discussing | racism. | graduate | discussing | racism. | education | discussing | racism. | considered | progressive | “diversity” | faculty | overcome | resistance | majority | colleagues | diversity | various | guarantee | racism. | openly | honestly | fragility | defensiveness | argumentation | certitude | natural | responses | attaining | productively | powerfully | hierarchy | ideologies | individualism | meritocracy | repetitive | media | representations | segregation | neighborhoods | depictions | whiteness | truncated | history | openly | solidarity. | interrupting | racism | ongoing | lifelong | conditioning | frameworks | simplistic | definition | racism—as | intentional | discrimination | committed | immoral | individuals | —engenders | confidence | evidence | implausible. | example | everyone | oneanother | discussion | engagement. | unconvincing | invalidate | experiences. | understand | socialization | UNDERSTAND | SOCIALIZATION |

    Fundamentals of Physics

    3+ syllable words: (show all ‘hard’ words)
    experimental | observations | quantitative | measurements. | objective | limited | fundamental | natural | phenomena | develop | experiments. | fundamental | developing | mathematics | experiment. | discrepancy | experiment | formulated | discrepancy. | satisfactory | limited | conditions | general | satisfactory | limitations. | example | discovered | century | accurately | comparable | relativity | developed | correctly | approaching | general | classical | developed | experiments | classical | mechanics | thermodynamics | electromagnetism. | important | contributions | classical | provided | developed | classical | mechanics | systematic | originators | calculus | mathematical | developments | mechanics | continued | century | thermodynamics | electricity | magnetism | developed | century | principally | apparatus | experiments | unavailable. | usually | century. | developed | discovery | physical | phenomena | classical | important | developments | relativity | mechanics. | relativity | revolutionized | traditional | energy | mechanics | microscopic | macroscopic | originally | formulated | distinguished | descriptions | physical | phenomena | atomic | constantly | improving | understanding | phenomena | fundamental | discoveries | overlap | chemistry | geology | biology | engineering. | notable | developments | numerous | astronauts | microcircuitry | computers | sophisticated | imaging | techniques | scientific | medicine. | developments | discoveries | society | discoveries | developments | exciting | challenging | benefit | humanity. | quantities | definition. | mechanics | quantities | quantities | mechanics | Measurements | measurement | measurement | meaningless | visitor | another | familiar | measurement | likewise | kilograms | kilogram | Whatever | readily | accessible | property | reliably—measurements | different | different | international | committee | established | quantities. | established | adaptation | abbreviation | International. | kilogram | respectively. | established | committee | temperature | electric | luminous | intensity | candela | mechanics | precisely | similarly | original | adopted | equator | particular | longitudinal | measuring | developed | scientific | everywhere. | recently | specific | platinum– | iridium | conditions | abandoned | several | principal | limited | accuracy | separation | determined | requirements | technology. | wavelengths | emitted | however | October | redefined | definition | establishes | precisely | approximate | assigning | numerical | various | physical | quantities | William | unsatisfactory | beginning | kilogram | specific | platinum–iridium | cylinder | International | Bureau | established | platinum–iridium | unusually | duplicate | cylinder | National | Institute | Technology | Gaithersburg | Maryland. | approximate | various | originally | rotation | however | therefore | defining | consequently | redefined | precision | obtainable | atomic | frequencies | associated | atomic | transitions | precision | equivalent | uncertainty | redefined | characteristic | frequency | particular | cesium | “reference | period | vibration | radiation | cesium | atomic | therefore | them—synchronized | necessary | Julius | calendar | discovery | measurement | intervals | interval | location | otherwise | example | positioning | satellites | unable | location | sufficient | accuracy | approximate | intervals | presented | addition | another | engineering | conventional | United | acceptance | National | Kilogram | accurate | International | Kilogram | National | Institute | Technology | primary | frequency | atomic | accuracy | courtesy | National | Institute | Technology | Department | respectively. | universally | accepted | industry. | limited | engineering | classical | mechanics. | addition | kilogram | millimeters | nanoseconds | various | frequently | various | abbreviations | example | equivalent | millimeter | corresponds | kilometer | likewise | megavolt | chemical | identity | indefinitely. | philosophers. | Leucippus | Democritus—could | speculated | ultimately | particle | atomos | sliceable. | ordinary | electrons | surrounding | following | discovery | particle | collection | particles. | composition | completely | understand | specifically | determined | occupying | entities | positive | specific | element | identified | atomic | element. | hydrogen | atomic | hydrogen | helium | atomic | uranium | atomic | addition | atomic | characterizing | atoms—mass | atomic | element | element | different | isotopes | another. | existence | verified | conclusively | primary | together. | repulsive | positively | particles | exotic | particles | different | varieties | particles | easily | likewise | density | property | density | usually | example | aluminum | density | density | therefore | aluminum | equivalent | densities | various | difference | density | aluminum | different | atomic | atomic | element | element | element’s | isotopes | relative | isotopes | relative | atomic | atomic | discrepancy | difference | atomic | separations | atomic | arrangements | relative | isotope | isotope | isotopes | different | practically | atomic | exactly | particles | molecules | particles | isotope. | particles | example | aluminum | experiments | Avogadro’s | Avogadro’s | exactly | general | element | element’s | atomic | element | element | Dimensional | Analysis | following | dimensional | analysis | determine | expression | relationship | dimensions | equation. | illustrate | procedure | formula | acceleration | expression | dimensional | analysis | validity | expression. | quantity | dimension | equation | dimensionally | quantity | dimension | dimensional | substituting | dimensions | acceleration | equation. | dimensional | equation | general | procedure | dimensional | analysis | expression | exponents | determined | indicates | proportionality. | relationship | dimensions | dimension | dimension | dimensions | acceleration | dimension | exponents | dimensional | equation | conditions | Returning | original | expression | expression | dimensionless | determining | dimensional | analysis. | conversion | necessary | another. | conversion | conventional | conversion | Appendix | algebraic | quantities | example | centimeters. | exactly | multiplying | multiplying | numerator | denominator | identical | estimates | OFMAGNITUDE | CALCULATIONS | approximate | physical | information | available. | approximate | determine | accurate | calculation | necessary. | approximations | usually | assumptions | modified | accuracy | magnitude | quantity | quantity. | example | quantity | magnitude | quantity | magnitude |

    • Replies: @Anonymous
  49. Emily W says:
    @res

    I would have expected having the IQ for both parents at follow-up to give a stronger correlation (opposite of what was seen). Similarly for the presumably higher IQ heritability in the older children. Any thoughts on that?

    The main reason for this is probably that the IQ test used at follow-up (ICAR-16) is much lower in reliability than the measures taken at intake (WISC/WAIS). It is unfortunate that the full Wechsler couldn’t be given to our follow-up participants, but we had to work with what we had. The ICAR-16 is great for what it is, but 16 items measuring four subcomponents can’t compare with a full WAIS.

    How comfortable are you with using the EA3 PGS for Asians given that it was developed with Europeans? I see significant effort to look at that in both the paper and the SM. Table S14 is the best summary of that I see. There we see the R^2 for PGS_EA reduced about 60% and 1/3 for Total IQ and ICAR-16 for Asians relative to whites. But the most interesting result is the Years of education R^2 for Asians being twice that of whites?! Any idea why that would be?

    It would be inappropriate to use these PGS for any sort of group comparisons, and the primary reason we included them was to test the hypothesis of the adoption placement effect. We were also puzzled by the seemingly high PGS estimates in a non-European sample too. But considering that mean IQ did not differ significantly across the groups we studied, the most probably reason for the unusually high estimate is allele ascertainment bias. Allele frequency ascertainment bias can arise from systematic deviations from the expected theoretical result due to sampling processes of the genotyping chip used to ascertain SNPs and their population-specific allele frequencies; a lower frequency of the minor allele in East Asians can cause the appearance of a higher PGS without underlying predictive significance.

    • Thanks: res
    • Replies: @res
  50. Emily W says:
    @res

    That is an extremely interesting point. Could you elaborate on that a bit? Will the presence of non-additive genetic variance always (or usually) lower C below its true value? With that variance ending up in E for an ACE model? Or would it appear in G-E covariance?

    Biometric models iteratively adjust variance components until they best match the constraints of equations describing familial relationships and the observed correlations. If the correlations between biological relatives is much higher than their degree of relatedness would suggest, you can end up with negative C. If you do a back-of-the-envelope heritability calculation with Falconer’s equations on twin data, you can see this effect in action: If the monozygotic twin correlation is greater than twice the dizygotic correlation, you’ll end up with negative C. So, e.g., if identical twins correlate on some phenotype at .7 and fraternal twins correlate at .3, you’ll have c^2 = 2(.3) – .7 = -.1. The logical consequence is that the presence of dominance variance will artificially lower C below its true value.

    We ran separate biometric models on our data which included dominance and excluded it, and the dominance estimate was consistently 0. Dropping the parameter from the model therefore produces better estimates of the remaining parameters.

    Incidentally, there is compelling theory and evidence for most genetic variance being additive (see, e.g., Hill et al., 2008 and Maki-Tanila & Hill, 2014) and dominance variance being negligible (Hivert et al., 2021 and Pazokitoroudi et al., 2021) in many psychological variables, including IQ and educational attainment. In fact, EA4 (currently in the works with >3 million participants) specifically ran a dominance model and found no evidence for dominance in educational attainment.

    I’m not aware off the top of my head of any human behavioral traits that consistently show evidence of dominance variance, but in general, traits that are under strong directional selection can sometimes show significant amounts of dominance relative to additive genetic variance. Such traits often include life-history phenotypes that are highly correlated with fitness and under strong selection pressure, particularly when studied in wild animals (see, e.g., Crnokrark & Roff, 1995).

    • Replies: @res
  51. res says:
    @Emily W

    the most probably reason for the unusually high estimate is allele ascertainment bias. Allele frequency ascertainment bias can arise from systematic deviations from the expected theoretical result due to sampling processes of the genotyping chip used to ascertain SNPs and their population-specific allele frequencies; a lower frequency of the minor allele in East Asians can cause the appearance of a higher PGS without underlying predictive significance.

    I’m not sure I understand how allele frequency ascertainment bias would inflate the variance explained for an Asian population using a chip (I think) designed more for Europeans as well as a GWAS done on Europeans. I thought those effects generally tended to decrease R^2 for PGS applied to other populations?

    If I understand correctly, one part of the problem is genotyping chips designed for one population might not pick up all relevant SNPs for another population with respect to that trait (a simple example is pop A fixed for a SNP which varies in pop B, but is not on the SNP chip). This reduces accuracy (and variance explained) for pop B because predictive SNPs for it might not be on the chip and/or found by the GWAS done on pop A.

    Another problem is differing linkage disequilibrium for the populations with respect to a given SNP. It seems given the right constellation of different causal variants in LD with the SNP for the two populations this could increase variance explained, but isn’t it more likely there would be a decrease?

    I know you and your coauthors have thought much harder about this from a stronger knowledge base. I am trying to improve my understanding. Thank you.

    For anyone else interested, here is a relevant paper.
    SNP ascertainment bias in population genetic analyses: Why it is important, and how to correct it
    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3849385/

  52. Rob says:
    @res

    Dr Thompson, you have probably seen this before

    @res Does that work here?

    I found this paper “Genetic enhancement of cognition in a kindred with cone–rod dystrophy due to RIMS1 mutation”

    They found a very rare genetic defect in RIMS1, a gene expressed in neural synapse. It’s also expressed in retina. The mutation seems to raise verbal iq, and the authors think g, by 25 points. The cost is going going blind.

    So, a very rare gene, phenotypic defect, and big boost to IQ. Sounds like organic genius to me. That’s just the kind of allele i postdicted. I feel smart now. We should look for more. If only to give kids who have them appropriate special ed. But really to figure out how to boost IQ without going blind.

    • Replies: @res
  53. Ron Unz says:

    Many of you are probably already aware of it, but a couple of weeks ago I added the MathJax technology to this website, allowing both articles and comments to render complex mathematical formulas, both inline and standalone, such as:

    $$\limsup_{n \to \infty} \frac{\pm S_n}{\sqrt{2n \log\log n}} = 1 \quad \text{a.s.}$$

    Here’s the MathJax website and documentation:

    https://www.mathjax.org/

    https://docs.mathjax.org/en/latest/basic/mathematics.html

    • Replies: @res
  54. res says:
    @Emily W

    Incidentally, there is compelling theory and evidence for most genetic variance being additive (see, e.g., Hill et al., 2008 and Maki-Tanila & Hill, 2014) and dominance variance being negligible (Hivert et al., 2021 and Pazokitoroudi et al., 2021) in many psychological variables, including IQ and educational attainment.

    Thank you! Especially all of the references. If anyone else is interested in digging in, I think these are the right links.

    Hill et al., 2008
    Data and Theory Point to Mainly Additive Genetic Variance for Complex Traits
    https://www.pure.ed.ac.uk/ws/portalfiles/portal/8168000/PLOS_2008_Hill_W.pdf

    Maki-Tanila & Hill, 2014
    Influence of Gene Interaction on Complex Trait Variation with Multilocus Models
    https://academic.oup.com/genetics/article/198/1/355/6073384

    Hivert et al., 2021
    Estimation of non-additive genetic variance in human complex traits from a large sample of unrelated individuals
    https://www.biorxiv.org/content/10.1101/2020.11.09.375501v1

    Supplementary Table 3: UKB Phenotypes used for analysis
    Lists the 70 traits analyzed. (no EA or IQ?)

    Pazokitoroudi et al., 2021
    Quantifying the contribution of dominance deviation effects to complex trait variation in biobank-scale data
    https://www.cell.com/ajhg/fulltext/S0002-9297(21)00102-6

    They looked at 50 UKBB traits, but I do not see EA or IQ.

    Much more from that group at http://web.cs.ucla.edu/~sriram/publications.html
    This seemed especially interesting.
    Efficient variance components analysis across millions of genomes
    https://www.nature.com/articles/s41467-020-17576-9

    Fig. 4: Estimates of genome-wide SNP heritability from RHE-mc, LDSC, S-LDSC, GRE, and SumHer for 22 complex traits and diseases in the UK Biobank.

    Does data exist in similar form for the cognitive variables (EA, IQ, ?) in the UKBB?

    In fact, EA4 (currently in the works with >3 million participants) specifically ran a dominance model and found no evidence for dominance in educational attainment.

    Thank you! Based on Steve Hsu’s work I had thought EA4 should be getting into the sample size range where you can detect nonlinear effects and I was wondering how that turned out. Was the model used testing for dominance at the level of individual SNPs or the overall effect?

    Steve talked about EA3 methodology a bit here.
    https://spartanideas.msu.edu/2018/07/31/ssgac-ea3-genomic-prediction-of-educational-attainment-and-related-cognitive-phenotypes/

    Although the study used over a million genotypes, the data had to be aggregated across many sub-cohorts using summary statistics only. This does not permit the L1-penalized optimization we used to build our height predictor.

    Does that also apply to EA4?

    On another topic, do you have any knowledge of large sample size GWAS work which looks at CNVs or other structural variation? Based on this schedule, whole genome sequences for 200k individuals in the UKBB should be imminent.
    https://www.ukbiobank.ac.uk/enable-your-research/about-our-data/genetic-data

  55. res says:
    @Rob

    @res Does that work here?

    Not sure what you mean.

    I found this paper “Genetic enhancement of cognition in a kindred with cone–rod dystrophy due to RIMS1 mutation”

    They found a very rare genetic defect in RIMS1, a gene expressed in neural synapse. It’s also expressed in retina. The mutation seems to raise verbal iq, and the authors think g, by 25 points. The cost is going going blind.

    Thanks. Though it is a small sample, Table 1 and Figure 1 seem rather convincing to me. There is a version of the paper with Supplementary Material (e.g. including visual acuity for affected individuals) at https://jmg.bmj.com/content/44/6/373

    Cited by information available at
    https://pubmed.ncbi.nlm.nih.gov/?linkname=pubmed_pubmed_citedin&from_uid=17237123
    Does not look like anyone followed up on the cognitive implications?

    It looks like they are singling out one mutation.

    In our kindred, RIMS1 mutation (Arg844His) causes a late-onset dominantly inherited cone–rod dystrophy (CORD7; OMIM 603649), leading to varying degrees of visual loss starting from the third decade onwards.

    This page gives it a different designation: arg820-to-his (R820H)
    https://omim.org/entry/606629

    This looks like the correct ClinVar page:
    https://www.ncbi.nlm.nih.gov/clinvar/variation/4168/
    which references this rsID:
    https://www.ncbi.nlm.nih.gov/snp/rs121918302?vertical_tab=true
    More at
    http://www.ensembl.org/Homo_sapiens/Variation/Summary?v=rs121918302;toggle_HGVS_names=open

    Seems odd that there has been no follow-up in the 14 years since that paper. Any further thoughts?

    • Replies: @Rob
  56. res says:
    @Ron Unz

    Thanks, Ron!

    Raches gave a brief explanation I found useful in this comment.
    https://www.unz.com/announcement/a-continuation-of-the-covid-vaxxing-debate/?showcomments#comment-4883838

    Write your maths in \(\LaTeX\), surrounded by \\(…\\) if embedded in the middle of a sentence—or surrounded by \\[…\\] before and after, to render “display maths” centered in their own block of text. This latter effect is seen in my prior comment. (There are other available delimiters involving dollar signs or double-dollar-signs, as Mr. Unz mentioned and I have used a few times.

    One nice feature of MathJax is it is easy to copy equations. Just right click on the equation and select either “Show Math As” or “Copy to Clipboard” and then “TeX Commands.” Here is Ron’s equation copied that way then surrounded by double dollar signs.

    $$
    \limsup_{n \to \infty} \frac{\pm S_n}{\sqrt{2n \log\log n}} = 1 \quad \text{a.s.}
    $$

    P.S. Any double backslashes are supposed to be single. Ron, for features like this it would help if preview worked better. In this case after editing the MathJax renders as text until the page is manually reloaded.

  57. Anonymous[939] • Disclaimer says:
    @res

    “The point was that CRT dullards love their “big” words to the point of feverishly inventing new ones on a daily basis. Their lectures and literature will be much denser, in more ways than one, when compared to Math or Physics studies.”

    Hard to argue with that ; )

    I’m glad we agree. This is important because later on, when comparing White Fragility and Fundamentals of Physics books, you seem to disagree.

    “A good IQ test should always put every test-taker on a level playing field in terms of knowledge.”

    I think you need a qualifier in that. Say “as much as possible”? You do realize as stated your goal is unattainable, right?

    Uh, please don’t do that. I don’t want to pepper my posts with tedious and unnecessary qualifiers. Let’s just say that nothing is perfect but some things are less perfect than others. In this case, my position is that verbal IQ is a much less perfect measure of actual intelligence than other IQ subscales. I am not one of those people who don’t “believe” in IQ tests – quite the contrary- but I do have a problem with one of the IQ subscales trying to measure intelligence by heavily measuring knowledge while all the others avoid it as a rule.

    “Verbal IQ is the only subscale that can be boosted through learning”

    That seems obviously wrong. Do you really think practicing Raven’s Progressive Matrices style questions would not help someone who was previously unexposed to that type of thing? At the extreme end, test prep which included the exact questions to be used later would presumably improve results on most tests.

    That’s true but you’re talking about actual cheating which makes your objection kind of a nitpick, again. I’m sure it happens in some cases, which is unfortunate, but the data I’ve seen so far doesn’t show a significant percentage of people increasing their non-verbal IQ scores over time. It’s a demanding task and test creators are obviously mitigating the problem with reasonable success.

    A person’s vocabulary, on the other hand, will diverge in different directions and at different speeds even without cheating. They’ll change their spouse, friends, work environment or hobbies and learn new words (or even end up shrinking their vocabulary in some cases). Change is actually unnecessary – an academic who communicates a lot with other academics will keep increasing his vocabulary regardless of any attempt at cheating. The same applies to bookworms or Grievance Studies adherents. They can be smart or dumb but they will expand their vocabulary – and verbal IQ – while their real intelligence (raw brain power) stays the same or even declines.

    Some dimensions to consider here.

    1. Variability of performance with learning. How much of a difference does the full range (say raised by wolves to reading voraciously since age 3) of learning variability make for IQ? How much of a difference does the typical range (say 5% to 95% of reading/learning effort in a population) of learning variability make for IQ? How much effort is required to move measured IQ? I would expect the improvement curve would be steep for much of going from raised by wolves to mandatory K-12 education, but much shallower after that.

    2. Persistence of learning improvements on measured IQ. I would expect specific training (e.g. tell people all of the words in Wordsum) to fall off quickly while general learning would be more durable (but also less dramatic in effect).

    Sounds reasonable. Please point me to any research dealing with those subjects.

    • Replies: @res
  58. Anonymous[939] • Disclaimer says:
    @res

    “Those are huge differences – and they have nothing to do with intelligence and everything to do with learning.”

    That false dichotomy is a gross overstatement. I’m curious, do you think intelligence affects the ability/efficiency of learning?

    Of course it does but “nothing to do with intelligence” in this case means that an average teenager will almost double his vocabulary by simply learning. College will push it from 12 000 to 17 000 words and then it will go to 17-21K or even much higher in some cases. They won’t become more intelligent in terms of raw brain power – which is what IQ tests should measure – but their verbal IQ will surge and the sky is the limit.

    Intelligent people will have higher “ability/efficiency of learning” but how is that relevant? They will also, on average, be much more inquisitive which will further boost their vocabulary-acquisition pace but that’s also irrelevant to my point? My point is that knowledge is not intelligence. That’s why we have school grades for knowledge and IQ tests for intelligence.

    “I’m willing to bet big money – without reading – that “White Fragility” contains many, many more “big” words than any Physics book of similar size.”

    I hope you haven’t been foolish enough to actually place that bet with anyone (especially given the “many, many more” bit). But if you’d like to do it now… I took roughly 3000 word excerpts from Chapter 1 of White Fragility and Fundamentals of Physics…

    I suspect that the sample size is too small. Something tells me that the anti-white crusader couldn’t resist adding more and more exotic words in subsequent chapters in order to appear smart while the physicist didn’t have that chip on his shoulder. I can’t be sure, though, so maybe you’re right but my experience with those two groups tells me otherwise.

    “tracking lifetime IQ measurements – showing verbal IQ trending upwards – up to a point – while every other IQ measurement and the average efficiency of physical brain cells declined. Again, one of these things is not like the others and that’s just laughably wrong.”

    The 389 page WJIV Technical Manual is an excellent reference for age curves of different subtests.
    https://www.wjscore.com/Files/WJIVTechnicalManual.PDF
    More at this comment.
    https://www.unz.com/jthompson/no-skin-in-the-game/?showcomments#comment-4100684

    See Figures 5-3 and 5-4 on page 137. You are right about one thing being not like the others.
    The Comprehension-Knowledge (Gc) factor cluster and Vocabulary (VL/LD) curves both show the gradual increase from about age 20 to 65 you mention. The others show varying levels of flat and declining after age 20. Cognitive Processing Speed (Gs) and Perceptual Speed (P) are notable outliers in the other direction (declining more quickly than the others).

    That PDF link doesn’t work, unfortunately, but I’ll try to find it through the data you provided when I get more time. Thanks. It does look like verbal IQ is “not like the others” which is part of my main point. A person can boost their score in that area even when physical brain deteriorates.

    “Bottom line is that data in this instance doesn’t matter. We can use reason.”

    Yes, we have often seen how well that works in the past. /sarc

    I was referring to that particular instance. It wasn’t a call to abandon facts in principle. I didn’t have a link to the study so I made an argument based on logic.

    • Replies: @res
  59. Rob says:
    @res

    Perhaps the going blind family did not want to be studied? I would not be leery of being studied for intelligence research, though no one’s breaking down my door to try. A quick googling and I could not find out who they are, names and locations, at least. Someone should preserve that, lest the researchers die without divulging.

    As to why there was no follow up, at least according to Google Scholar, I would guess that researchers are leery of doing research that smacks of eugenics. There does not even have to be an official policy. Just don’t give grants for that. Or make getting ethical clearance nearly impossible. I’ve read that Rushton was a pariah at his uni, no grad students for decades, office in the basement, that sort of thing,

    Consider the comparison to doing genetic studies of American Indians. It’s not illegal, but no researcher wants to be at the center of a controversy, so they don’t even try. The promise of tenure selects for people who value job security. That they have tenure at one school means they don’t want friction with colleagues.

    Speculation aside, talking to the reasearchers is probably the best way. There was a corresponding author on the RIMS1 paper, you might consider emailing him about why they didn’t follow up. I might, too.

    I would like to know how that family was discovered, and why doctors thought they were brainy. It’d be a hint on how to find more organic geniuses. I think they’ll be reluctant to come forward, because disabled and different kids are mistreated by schools and other kids. It leaves scars.

    I’d like to toot my own horn and say my method of looking for siblings with big IQ Δ would have found these people. Also my method of looking for very smart people with physical (I don’t want to say defects) differences and disabilities.

    Twenty five points is a lot 1.67 σ. Difference between retarded and Normal. Between normal and solid college material. Between that and genius. Or even between genius and your name will never be forgotten. And the ceiling is above 150! Two people in a thousand for a normal population.

    You know, there are lots of families in which all the kids take the SAT, though in different years. College board probably has records of families where one got 1550 and another 900. That would be interesting to have. Does it happen more often than one would predict from the σ of 12 within sibships?

    My idea of organic genius is very different from folk theories of smart people going crazy. I concede that most eg schizophrenics have low IQ. I’m saying that some don’t, and they’d be interesting to sequence. An analogy would be the analysis that “proved” PANDAS (pediatric acquired neurological disorders associated with strep) by looking at all the kids who got strep, But strep causes crazy was not really the claim. The claim was “some strain of strep causes crazy” they would have to sequence the streptococcus bacteria from all these kids to test it. That said, not sure I believe in PANDAS.

    Looking at ways to make dumb people smarter is probably ok with their peer group like I read a piece about a doctor/scientist (i forget) who is working on a way to make people with Downs more intelligent. The drug is called mementine, I think. On the other hand, the young college-educated set is extremely woke.

    Remember Greg Cochran’s thing about heterzygosity for some sphingolipid disorders boostin IQ? I wonder if youncould sell a pregnancy vitamin pill with iodine and some sphingolipids or bprecursors to pregnant women. Or dorks who want to get smarter. Add creatine, too. Here’s a review of creatine in healthy people. Some evidence it works, especially in vegetarians. Sell it to Indian students (the kind who live in India)

    What works on people with trisomy 21 might not work in normal people, but my guess is it would. Teenage boys produce a ton of testosterone, but steroids work for them.

    One could write interesting sci-fi about engineering or breeding nearly superhuman geniuses who go blind in their forties. Along the lines of Camp Concentration, maybe. When is it ethical to create people who will be severely disabled at a young age? That it’s based on a real condition makes the sci-fi hard. Maybe someone invents usable synthetic eyes, and the geniuses become much harder to control. Give the silver eyes in the movie, maybe opalescent,

    Sorry, got distracted.

    I would like to see how the RIMS1 kinship bunch do when their eyes have not declined and they can take a full IQ test. Are they even smart as kids? As babies or toddlers?

    I’d like to see this mutation put at the appropriate place in the mouse RIMS1. Also macaques or chimps. Would like to see a homozygotes- do they get 50 points?

    It is interesting that this intelligence-enhancing gene affects the eyes. Coupled with the fact that near-sighted people are more intelligent, i envision that eye differences might be a fruitful place to look for organic genius. Like blue eyes. They are so common in some populations, but universal in none. Doesn’t that imply heterozygous advantage? Further, the allele is not old and the haplotype is long. It was under positive selection in heterozygotes. Maybe the blue eyes genes affect cognitive ability or profile? On Twitter, Uriah(@crimakid, and apologies if you led me to him) noted that NFL quarterbacks who get to the super bowel are bluer eyed than NFL QBs, who are themselves more likely to have blue eyes.

    Back to RIMS1, I would love to know what it does in the brain and when it matters for intelligence. Worse comes to worst if you’re genetically engineering embryos, put a standard RIMS1 gene under a promoter for retina-only genes, and RIMS1-IQ under a brain-only promoter. Add a microRNA for the mutant gene in the retina to reduce off-target expression. Viola! smart kids who don’t lose their sight.

    Off this gene. It is shocking how little serious research is done on atrempting to understand and increase intelligence. Considering that raising IQ 25 points across the board would be utterly transformative. Every serious problem, from climate change to dempgraphic realities would be transformed. We’d get new problems, of course. If we could do it, and China could not, it would ensure our status as the hyperpower, despite our unfortunate demographic choices. Vice versa, and no one ever catches China. They’re going to the stars.

    If IQ went up nearly 2σ across the board, then we would solve real problems scientifically instead of dealing with the results politically. We would cure aging, not fiddle with social security. We’d create a von Neumann economy, where machines reproduce themselves towards human ends, rather than fiddle with interest rates. We’d prevent hurricanes, instead of building higher levees around New Orleans.

    Perhaps China getting a strong head start on upping IQ would get us in a space race sort of mindset? Except for satellites, space is not that important. Well, place ICBMs go through, too. But intelligence affects everything.

    I’m sorry this is not a papers-focused comment, but it is late.

    • Replies: @res
  60. dearieme says:

    in the present sample, who are predominantly international placements with limited information on birth background …

    Gee whizz, how many complications might that produce?

    • Replies: @dearieme
  61. dearieme says:
    @dearieme

    Hold on. How is that compatible with Study eligibility was limited to those families living within driving distance of the research lab ?

    This paper is too ill-written for me to take it seriously. They simply don’t explain their vocabulary unambiguously. Presumably I am meant somehow to intuit the meaning of their sloppy English. The referees should be shot.

  62. res says:
    @Rob

    Perhaps the going blind family did not want to be studied?

    Sure, but the most interesting follow-up would be to look at the SNP (I think it can be reduced to a SNP) in other contexts. The small sample size is an issue (and I was bothered by how many of the non-affected were missing data).

    As to why there was no follow up, at least according to Google Scholar, I would guess that researchers are leery of doing research that smacks of eugenics.

    Probably, but if there really is a SNP that boosts IQ that much I would expect someone to be interested. But then I would say the same thing about the DUF1220/Olduvai domain CNV, but have not seen much since this 2015 paper.
    DUF1220 copy number is linearly associated with increased cognitive function as measured by total IQ and mathematical aptitude scores
    https://pubmed.ncbi.nlm.nih.gov/25287832/

    15 citations and the only one which looks related to intelligence is this 2017 paper
    A Genome-Wide Association Study Identifies Genetic Variants Associated with Mathematics Ability
    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5290743/
    Which simply notes: “some studies reported sporadic association of copy number variations with mathematics ability but few of them been validated independently.”
    (also notice that the authors are from Shenzhen, China)

    I think your sibling IQ delta method would be worthwhile, but you would probably need to sort through many examples of simple defects and some good/bad pairs of rolls of the additive dice (when you are looking for one in a million effects low frequency confounders become an issue). For something that infrequent you would really need a large sample (SATs seem like a good idea). Given all of this, your questions about how this were found are interesting. I’d like to hear if you do ask the researchers.

    Sibling methods in GWAS seem to be becoming more popular. From 2021.
    Within-sibship GWAS improve estimates of direct genetic effects
    https://www.biorxiv.org/content/10.1101/2021.03.05.433935v1.full

    From 2020.
    Sibling validation of polygenic risk scores and complex trait prediction
    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7411027/

    I think the UK Biobank including a significant number of siblings pairs
    https://www.researchgate.net/figure/Sample-sizes-for-different-family-relationships-in-the-UK-Biobank_tbl1_331911415
    is one driver of this.

    Remember Greg Cochran’s thing about heterzygosity for some sphingolipid disorders boostin IQ?

    I remember. Something else which I am surprised by lack of follow-up. I expected relevant SNPs to show up in GWAS. Would be interesting to try to understand why not (no effect? too rare? close but no cigar p value?).

    For RIMS1 I think it would be better to try to replicate the effect in another sample before delving into trying to understand it. My guess is things are more complex and it would be challenging to replicate with anything like the same effect size. This is the type of thing where publication bias can be a problem.
    Publication bias – Importance of studies with negative results!
    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6573059/

    One could write interesting sci-fi about engineering or breeding nearly superhuman geniuses who go blind in their forties.

    I assume you have read Flowers for Algernon? Your premise is intriguing–I leap to the in/voluntary aspect.

    If IQ went up nearly 2σ across the board

    Do you read Steve Hsu’s blog? If not, you should search his archives. The tags in this post would make a good start.
    https://infoproc.blogspot.com/2014/10/big-chickens.html

    Same for Gwern. Good starting point there.
    https://www.gwern.net/Embryo-selection

    If you know of similarly interesting sites on that topic which I have not mentioned, please let ME know ; )

    • Replies: @Rob
    , @Factorize
  63. res says:
    @Anonymous

    My point is that knowledge is not intelligence. That’s why we have school grades for knowledge and IQ tests for intelligence.

    But they correlate quite well. Which can be useful. Educational Attainment correlates even less than vocabulary (or grades) with intelligence (I believe) and we still find it useful as a proxy.
    From this page WORDSUM correlates 0.71 with adult IQ.
    https://emilkirkegaard.dk/en/2019/01/the-wordsum-question
    And Table 1 there shows Education correlating 0.66 with adult IQ and 0.51 with WORDSUM.

    The current paper (Table S8) gives a correlation of 0.35 between Years of Education and Total IQ. Has the correlation really gone down that much since the 1940s data at Emil’s page? Seems possible given how higher education was so much less common then.

    I suspect that the sample size is too small. Something tells me that the anti-white crusader couldn’t resist adding more and more exotic words in subsequent chapters in order to appear smart while the physicist didn’t have that chip on his shoulder. I can’t be sure, though, so maybe you’re right but my experience with those two groups tells me otherwise.

    Just so stories are easy to make up. Feel free to try the experiment yourself on later chapters. Do you think physics textbooks become less complicated in later chapters? I would argue that trend is more pronounced there than for popular books

    Did you look at the 3+ syllable word lists? I thought those from White Fragility were markedly less sophisticated. Might be worth trying to quantify both that and repetition, but I am not sufficiently motivated.

    That PDF link doesn’t work, unfortunately

    Sorry about that. I thought I checked, but must have just searched for the file on my HD after I saw the name in the link in my old comment. BTW, searching the internet for the filename is a good way to find things which have disappeared. In this case the file is available at:
    https://pdf4pro.com/cdn/1588280-wjiv-techman-cd-insert-wjscore-com-59b916.pdf
    with an archive version at
    https://web.archive.org/web/20210511182047/https://pdf4pro.com/cdn/1588280-wjiv-techman-cd-insert-wjscore-com-59b916.pdf
    in case that disappears.

    I was referring to that particular instance. It wasn’t a call to abandon facts in principle. I didn’t have a link to the study so I made an argument based on logic.

    Fair enough, but I think my point gives good reason to be skeptical. Also worth noting that “data…doesn’t matter” was a bit strong. It sounds like you meant more like “is not required.”

    I think a big part of our disagreements is you like hyperbole. I find it fun for humor, but not so much for thoughtful conversation.

  64. res says:
    @Anonymous

    I’m glad we agree. This is important because later on, when comparing White Fragility and Fundamentals of Physics books, you seem to disagree.

    I was more responding to the overall point and humor. Regarding the simple interpretation (denser = more technically difficult) of “Their lectures and literature will be much denser, in more ways than one, when compared to Math or Physics studies.” I do disagree with that. Mostly because the difficulty is of fundamentally different kind.

    Physics is trying to understand concepts which may be difficult (especially if one does not like math), but tend to unify as they become better understood.

    Grievance Studies tends to be about memorizing inconsistent points presented as obscurely as possible. If anything, I think the better you understand it the less comprehensible and more ridiculous it becomes. I believe it loads much more heavily on memory (and willingness to ignore contradiction) as well as verbal fireworks.

    Uh, please don’t do that. I don’t want to pepper my posts with tedious and unnecessary qualifiers. Let’s just say that nothing is perfect but some things are less perfect than others. In this case, my position is that verbal IQ is a much less perfect measure of actual intelligence than other IQ subscales.

    Finding the right balance of qualification can be difficult (I think we tend to err in opposite directions, part of the problem here). Consider your quote. How about just removing the adverb “always”–which is what I see as the problem.

    “A good IQ test should always put every test-taker on a level playing field in terms of knowledge.”

    Both shorter and more accurate IMHO.

    In this case, my position is that verbal IQ is a much less perfect measure of actual intelligence than other IQ subscales.

    See Table S8 of the adoption paper. Correlations with Total IQ.

    1. Highest degree .32
    2. Years of education .35
    3. SES .23
    4. PGSEA .27
    5. ICAR-16 .43
    6. Verbal IQ .86
    7. Performance IQ .81
    9. Information .78
    10. Vocabulary .77
    11. Picture completion .63
    12. Block design .66

    I think that argues against your assertion. BTW, that ICAR-16 correlation with Total IQ is depressingly low.

    Table E-5 on page 313 of the WJIV Technical Manual has Test Score Intercorrelations—Ages 20 Through 39
    Table F-5 on page 319 of the WJIV Technical Manual has Cluster Score Intercorrelations—Ages 20 Through 39.

    The former has Vocabulary, and the latter General Intellectual Ability but I can’t look at the two together. Some good information in general, but nothing I see definitive for your point.

    That’s true but you’re talking about actual cheating which makes your objection kind of a nitpick, again.

    The first example was something which is sometimes included in test administration. The second was an extreme (I even used that word) example to emphasize the point. I would call your response more the nitpick since I think that form of hyperbole to imply the possible range of interventions is useful. Test prep sometimes gets quite far into that gray area.

    The same applies to bookworms or Grievance Studies adherents. They can be smart or dumb but they will expand their vocabulary – and verbal IQ – while their real intelligence (raw brain power) stays the same or even declines.

    But the more intelligent will be both more inclined to do that learning and more effective at it. Hence we end up with a useful correlation.

    Sounds reasonable. Please point me to any research dealing with those subjects.

    Unfortunately I don’t have any directly applicable at hand. If anyone does, please post.

    1. is hard and I don’t know if such research even exists.

    2. Lines up reasonable well with research on early childhood interventions so maybe that is of interest. This meta study seems like a good starting point.
    The environment in raising early intelligence: A meta-analysis of the fadeout effect
    https://www.researchgate.net/publication/284123208_The_environment_in_raising_early_intelligence_A_meta-analysis_of_the_fadeout_effect

    BTW, a naive look at that figure makes me wonder if some of those interventions really are effective. Hard to be sure with all of the noise, but I am not sure it is reasonable to lump such different programs together. Table 1 has a nice summary of the studies.

    • Replies: @Anonymous
  65. Anonymous[209] • Disclaimer says:
    @res

    Thanks for the links. I did use that method to quickly look for the PDF but that resulted in a document dealing with disabilities and a number of registration requests. I hate it when they do that. Anyway, I’ll take a look when I get the chance.

    In the meantime, I stand by my conviction that the verbal IQ component is deeply flawed. It measures vocabulary – which is a form of knowledge and an automatic red flag in my opinion. Every other IQ subscale avoids measuring knowledge for a good reason. Call me a purist, but if I wanted to hire genuinely intelligent people, who excel at problem-solving and unique insights, I’d use IQ tests without verbal components or ignore verbal components in those that do.

    I keep hearing that verbal IQ is useful. That’s clearly true. I never said it wasn’t. I just believe that it should be separated form tests that are supposed to measure raw intelligence or heavily modified to minimise measuring knowledge. It’s not gonna happen, of course, but that’s fine – I’ll just ignore it or use it to locate communicators.

    I also keep hearing that vocabulary and verbal IQ correlate well with intelligence. True, again, but it also correlates well with CRT mediocrities, introvert bookworms and all kinds of random environmental and life-choice factors that might not require high ability. Furthermore, the fact that it does correlate well with intelligence creates another problem that proves my points. Intelligent people will continue increasing their vocabulary (on average, of course) at a much steeper rate than most others all the way into the old age. Their education, work and social environment will boost it and the fact that they’re more inquisitive will often result in a lifetime of informal studying.

    That’s great, of course, but it creates a problem with measuring IQ. Whatever verbal IQ gap they had at 20 in comparison to an average person will keep widening.

    This leads me to the second red flag: improved verbal IQ scores as the brain ages and deteriorates. Thank you, again, for linking to a study that actually shows that it’s happening but everyone is avoiding talking about it. My first question when noticing the problem was WTF but the next two were more useful: why and what does that mean? The answers are “because it measures knowledge” and “it’s flawed”.

    • Replies: @res
  66. Rob says:
    @res

    You know, it is quite possible that someones are interested in IQ booster genes. There may be startups working with, just off the top of my head, RIMS1 variants in organoid brainlets. I bet it’s possible to do an in vitro “eyelet” with enough retina to see RIMS1 variants doing their thing. These companies don’t talk about it. They don’t want controversy, like people screaming about eugenics and the Holocaust or the poor being left behind as the rich become genetic supermen.

    But when they actually have a product along the lines of a couple does the in vitro thing and gets a dozen viable zygotes. then the company applies proprietary treatment x, which will cost, I dunno, 100k at first. Out of the dozen, two are viable and modified, normal RIMS1 in the eye, fancy RIMS1 in the brain. At that point, when they go into clinical trials, the screaming starts, but it’s muted by the little voice in the critic’s head, ‘my baby could be way smarter than other people’s.” Liberal critics can even make fun of conservatives who approve of tech developed with fetal tissue, aborted naturally or not. Religious conservatives might not approve, but they have influence through Republicans, who never win. Once a seriously effective, voluntary “eugenic” procedure is actually possible, then opposition becomes much more pragmatically focused.

    That would explain why there has been no published research. Companies do not have to announce publicly what they want to do, and technically potential investors are bound by non-disclosure.

    [MORE]

    On the sphingolipid disorders, he’s right about there being too many in one aspect of neural physiology to be chance. Has no one thought to feed sphingolipids to mice? Or transfusing mice. On GWAS, I looked for one on a condition with a known monogenic cause. Someone did a GWAS for sickle cell anemia. The known hemoglobin mutation fell out as extremely strongly associated. Another allele of a different gene was a hit, too.

    Has anyone done a GWAS on Ashkenazi Jews and intelligence/cognitive ability/Edu attainment/similar? This is where not knowing the math gets me. If Jews are a small percentage of a mixed ethnicity, white “amerimutt” sample population, would the effect of almost exclusively Jewish alleles be masked out from whatever is done with variants associated with principle components? Like, Jews are around half European and evolved under different, intense selection pressures with next to no gene flow after the early admixture. They are less similar to, say Dutch and Swedes than the former two ethnicities are from each other. Honestly, if middle easterners and mulattos are not included in GWAS looking at European/white Americans, probably Ashkenazi should not, either. This is an honest question. I know you are very bright, and you have a Gold star from Ron Unz, but I don’t know where you fall on the sane person to raving antisemite in Sailer’s comment section. On second thought, hmm a yellow star?

    I saw a GWAS-derived PGS applied to (sub?)sample of Ashkenazi from either that study or separate, and the Jewish people had high PGS scores for IQ or edu, but if I recall, the PGS was more predictive for whites. To me, that seems like there are alleles, either common or rare, that cause variation in the Jewish population that are not present in generic whites.

    Someone should do a full Ashkenazi-only GWAS while there are still some full Jews who are not extremely, weirdly (no judgment!) religious. Maybe Israeli researchers have done that?

    An interesting study to do would be to look at half Ashkenazi half generic American siblings. Take fraternal twin pairs, is the average within-pair ΔIQ larger for pairs that are discordant for the sphingo alleles than for pairs that both inherited the allele or both did not.

    What brought Jews to mind was not GWAS, but the culture of biomed researchers. Like any area where high IQ people thrive, Jews are over-represented in the field, and more over-represented the higher you go. As far as sciences go, bio and biomed (biochem less so) are heavily female. Becoming a researcher takes a lot of education, so they live in university environments for 10 years, say 18-28. That’s when adult culture gels in people. They are extremely university progressive. They believe in critical race theory. They think that sure the white population is declining, but Hispanics and blacks will become R&D producers who are just as good. Because the one black guy in their biomed grad school program was not the worst student. They do not have enough experience with us selected blacks, or even unselected whites, to have much inkling that other people have the culture produced by the personality types in that culture interacting. Because biomed requires learning a lot, and hard work from anyone who wants to do well in safe, drudgery-filled areas of science, they think they got their through hard work.

    Tl;dr the women Jews and gays produce a culture that is extremely hostile to bio-realism, much less race realism. They are not well-read outside of what schools assign. They think genetics/evolutionary theory applied to people = eugenics = mass murder. It does not help that every bio-realist place on the web accumulates people who think genocide should be on the table as a solution to existing problems. They do, however, think they are smarter than uneducated whites. David Barash, a psychologist bravely applying evolutionary theory to sex differences in sexual interests, said nasty things about the sigmatists/sigmastics. Barash is one of the evolutionary psychologists surveying college freshman psychology classes to learn deep truths about human evolution- which stopped at the neck the day the first person stepped out of Africa.

    Hey found Barash’s letter it’s the last one. It is about this article by Amy Harmon, I think it’s her first foray into IQ and biology.

    People don’t get grants to study things that make people smarter. Especially since Half-Sigma commenter Marc put IQ and schizophrenia-influencing alleles into the HapMap web interface and discovered that there were racial differences in the alleles’ frequencies. Amy Harmon (Horny Mama, from a non-Reg-Cæsar anagrammer) wrote a New York Times article about it. David Barash

    The sci-if idea? I was kind of thinking of Ender’s Game, where the government was desperate for genius kids. If you need a brilliant general in twelve years, then could use him for ten, and he goes blind at forty, then the temptation is there. So maybe the government does not need a general. Maybe they need brilliant people to fight global warming. Maybe someone seriously smart makes it into government and made this as a blue sky program to improve everything.

    They have made the first round of embryos from sperm and egg donors who are themselves very smart. The program is leaked, and the discussion of “should they do it” becomes public and very heated. While they are doing it, then that could be an interesting book or movie. You can have a President who is iffy on the thing but inherited the program from the former guy. His constituents hate global warming, poverty, disease, etc but they also hate inequality, IQ… they are progs. Can he build a new constituency to solve the most important problem the country faces: he could lose reelection.

    The beautiful scientist with the blind daughter. The daughter is blind Greta, really hates global warming. Daughter can be girl Tiresias, making predictions that turn out true but twisted, or Cassandra, true prophecy, but never heeded – this is a literary device, not fantasy crossover. How could she have worked on a program to make people who go blind? (she was on the very technical end and thought she was developing eye organoids to cure blindness, not so kids could get a fifty-point boost and not go blind until 30.

    The single, not very intelligent women who are going to be the surrogates. Some just want to have bright babies who are not as disappointing as their natural ones.

    The FBI agents tasked with protecting the surrogates, as right-wing terrorists target them for reasons. Gotta have this as pro-progressive, otherwise, it would never be published.

    The desperate owner of the company who developed the genetic modification tech. The kids are already going to be born. His company is obsolete. Can he convince the FDA to let a second-generation treatment or other gene mods go into trials before we know the results of the RIMS kids, in forty years?…

    You could set it in a European dystopia, like Transhuman Space, where neoliberalism reigned, tech advanced, and there were no disasters, yet. Or set ten years from now. Or tomorrow.

    The people opposed could be midwits in their early twenties: people who will do well fairly easily by being smarter, but realize they will be far outclassed by the next generation. From 95th percentile to average. Or worse, from 95th percentile to 5th percentile. Totally outclassed. Would cause resentment.

    The next book, after kids are born, maybe when they’re 15. Oh, the whole thing could be YA Fiction. I have a friend who wrote a YA book, maybe I’ll talk to him.

    I’ve read both Hsu and Gwern occasionally. Would like Gwern more if he wrote more like a blog, and Hsu if he wrote more like Gwern, They are both extremely bright. I hope Gwern has a well-paying job he loves, and pretty sure Hsu does. But take Hsu’s PGS for embryos company, he’s looking for something like a five-point IQ advantage over rolling the dice. That’s not much? I think engineering has to be the way to go, in the long run. Consider spell-checking, in a way it is very conservative, you are just eliminating extremely rare alleles, pretty much all neutral or harmful, but you eliminate the raw material of evolution. Not to mention, if you can spell-check an embryo, then you can do pretty much anything. That would be hundreds of edits. You’d probably get a large boost, though. For parents, some of those rare variants are what makes them, well them. Spell-checked children would be very generic, even if you left all the common 1% or higher differences.

    Do you know if they can do embryonic stem cells to haploid egg derived from the stem cell without going through the whole “nine months of pregnancy to get a woman to get the egg from in a decade or more? I have a business idea if that is possible. I think it’s a good one.

    I’m super-into intelligence and IQ because I have a 40-something gap between verbal and performance IQ. The shrink who tested me said it was the largest gap he’d ever seen in an adult who was not impaired. So I can see both the very smart world and the just trying to get out of this class without a D world. Though that wax before my back surgery, I might have stronger working memory now.

    I don’t think she’s as respectable as those two, but Pumpkin Person is knowledgeable about IQ, though not genetics.

    Honestly, the biology of intelligence is kind of a “when guns ate outlawed, only outlaws will have guns” situation.” “when biology is considered evil, only evil people will have biology.” Like, if you are a totally normal person, however bright you are, there is little upside to being thought of as thinking intelligence is biological and mostly immalleable with the technology we have today. Ergo, only people who either cannot fit in or people looking to pick up the trillion-dollar bill on the sidewalk are into this.

    Yet again, it is late. I will give some thought to Hsu/Gwern quality places.

    • Replies: @res
  67. res says:
    @Anonymous

    Call me a purist, but if I wanted to hire genuinely intelligent people, who excel at problem-solving and unique insights, I’d use IQ tests without verbal components or ignore verbal components in those that do.

    I’m somewhat sympathetic to that view (e.g. see my occasional rants about high-V low-M people and their leading role in our public society, contrast China with a significant number of leaders with a technical background), but I think you are underestimating the value of verbal IQ in creating a well rounded intellectually creative person. Even in STEM.

    Anne Roe’s book The Making of a Scientist has intellectual profiles of top scientists in the 1950s.

    Worth a look to see how high the verbal scores were for most of them. PDF at
    https://www.gwern.net/docs/iq/roe/1953-roe-makingscientist.pdf

    The SMPY has M/V/S profiles for their subjects and is worth a look to get a sense how those affect outcomes. The 35 year follow-up is a good starting point. Followed by a link to the immense related body of research.
    https://my.vanderbilt.edu/smpy/files/2013/02/DoingPsychScience2006.pdf
    https://my.vanderbilt.edu/smpy/

    That’s great, of course, but it creates a problem with measuring IQ. Whatever verbal IQ gap they had at 20 in comparison to an average person will keep widening.

    The question becomes which are the more accurate measures. Note that the heritability of IQ increases with age.

    This seems like a good time for the George Box aphorism: “All models are wrong, but some are useful.”
    https://en.wikipedia.org/wiki/All_models_are_wrong

  68. res says:
    @Rob

    Good point about unpublicized research. Would be a risky strategy for companies though. Too easy for others to discover your trade secrets. Though if everyone else is scared off of doing the research it might work.

    Has anyone done a GWAS on Ashkenazi Jews and intelligence/cognitive ability/Edu attainment/similar? This is where not knowing the math gets me. If Jews are a small percentage of a mixed ethnicity, white “amerimutt” sample population, would the effect of almost exclusively Jewish alleles be masked out from whatever is done with variants associated with principle components? Like, Jews are around half European and evolved under different, intense selection pressures with next to no gene flow after the early admixture. They are less similar to, say Dutch and Swedes than the former two ethnicities are from each other. Honestly, if middle easterners and mulattos are not included in GWAS looking at European/white Americans, probably Ashkenazi should not, either. This is an honest question. I know you are very bright, and you have a Gold star from Ron Unz, but I don’t know where you fall on the sane person to raving antisemite in Sailer’s comment section. On second thought, hmm a yellow star?

    I have posted over a million words of comments on the Unz Review so a search of my comments should give a pretty good sense of where I stand on different topics. Let’s just say I take fire from both sides in the JQ wars over in iSteve. This comment (and the associated thread) will probably tell you what you want to know:
    https://www.unz.com/anepigone/the-marriage-pot/#comment-4742924

    I don’t know of any Jewish-specific GWAS and suspect you are right about the PC corrections obscuring Ashkenazi-specific variation. I still would have thought there would have been enough signal to get through though if the effect is as large as posited.

    Here are two papers using PGS to look at Jews. From your comment below I’m pretty sure you have read the first, but not sure about the second (see Figure 9).
    https://www.researchgate.net/publication/330601752_Polygenic_Scores_Mediate_the_Jewish_Phenotypic_Advantage_in_Educational_Attainment_and_Cognitive_Ability_Compared_With_Catholics_and_Lutherans
    https://www.mdpi.com/2624-8611/1/1/5/htm

    If you know of any others please let me know.

    Also, the small proportion of Jews in most populations will limit the % variance explained of Ashkenazi-specific variants and make them harder to detect.

    I saw a GWAS-derived PGS applied to (sub?)sample of Ashkenazi from either that study or separate, and the Jewish people had high PGS scores for IQ or edu, but if I recall, the PGS was more predictive for whites. To me, that seems like there are alleles, either common or rare, that cause variation in the Jewish population that are not present in generic whites.

    That’s pretty much what you would expect for a GWAS developed for Europeans.

    But take Hsu’s PGS for embryos company, he’s looking for something like a five-point IQ advantage over rolling the dice. That’s not much? I think engineering has to be the way to go, in the long run.

    See the link I gave earlier:
    https://www.gwern.net/Embryo-selection
    and search for “multi-stage” or “iterated”.

    I don’t think she’s as respectable as those two, but Pumpkin Person is knowledgeable about IQ, though not genetics.

    PP’s blog is interesting and worth a look, but tends not to be as rigorous as Steve and Gwern. Also, the commenters…
    Just took a quick look and the posts seem better and the worst of the commenters not as vocal. Perhaps I should start spending more time there again.

    • Replies: @Rob
  69. @res

    at the very very top, the great scientists all had out of this world V scores.

    Gauss vacillated between being a philologist and a mathematician, Kolmogorov as well.
    Agassiz and Darwin had near-genius prose styles.
    Newton is often mocked for his studies of Hebrew texts, but his Bible studies are not mocked by people who know what they are talking about.
    Feynman was reputed to be one of the greatest conversationalists of his day – almost everyone who talked with him one on one agrees, except for a few people who had their own agenda.

    A counter-example is von Neumann, but he just didn’t ever bother to try. Also, most people who remember talking to him did not talk to him in Hungarian.

    Among the greats I have not mentioned (in other words, among the 20 or so most famous scientists of the last hundred years), about half clearly suffered from some form of autism.

    And when I say suffer, I mean suffer, in every sense of the term. My guess is that the autism knocked down their actual V scores by several standard deviations, with no benefit to their scientific insights – most people don’t think of it that way, but imagine, if you will, a Godel or a Dirac who were normal guys.

  70. Anonymous[956] • Disclaimer says:
    @res

    …but I think you are underestimating the value of verbal IQ in creating a well rounded intellectually creative person. Even in STEM.

    Why would you think that? Maybe you’ve decided that I have some kind of chip on my shoulder and that leads you to make these unfounded guesses based on my supposed motives. That’s a fool’s errand. My motives are irrelevant. I did score exceptionally well on two official IQ tests (in my native European language – not English) and one unofficial, non-verbal IQ test later on. Verbal IQ scores weren’t standing out in any direction, so when you’re responding to that imaginary person it only distracts from what’s written on the screen.

    That said, in your case it’s more likely that you’re being a pedant, again, and latching on the fact that I’ve written “ignore verbal IQ” without any qualifiers so now I’ll end up hiring a bunch of deaf-mutes from Timbuktu. I don’t think so. What I’ll avoid with my approach is hiring mediocrities who’ve stumbled into solving crossword puzzles because their father liked them or bookworms who ended up there because of their shyness. Only in Hollywood movies is that scrawny, bespectacled kid an automatic genius with a time-machine in the basement.

    As for “well-rounded”, I think they’ll be sufficiently well-rounded with Thompson’s 17-21K words (plus the specialised jargon in their field) and if they ever double it their verbal IQ would skyrocket while boosting their actual intelligence by… 0%.

    Worth a look to see how high the verbal scores were for most of them.

    How is that relevant? Did I ever say that legitimately intelligent people have deficient vocabulary?

    The question becomes which are the more accurate measures.

    Top speed, acceleration and horsepower are excellent measures of car performance. They will also, predictably, decrease with time. Mileage is not, but feel free to inform the owner about the good news. One of the measurements is going up, after all, and there’s certainly some positive correlation between good cars and how much people like to drive them.

    • Replies: @res
  71. res says:
    @Anonymous

    Call me a purist, but if I wanted to hire genuinely intelligent people, who excel at problem-solving and unique insights, I’d use IQ tests without verbal components or ignore verbal components in those that do.

    So you are saying I should not take that at face value? Seemed pretty clear to me.

    certainly some positive correlation between good cars and how much people like to drive them.

    I am very much enjoying your apparent inability to make reasonable relative comparisons (maybe that is why you keep taking refuge in hyperbole and false dichotomies?). We aren’t talking about “some positive correlation” here. Note that the Verbal IQ correlation with total IQ was 0.86 which was higher than the Performance IQ correlation with total IQ of 0.81.

    What is fascinating about this conversation is how you extol the virtue of non-verbal IQ in preference to verbal IQ yet you completely ignore the data I present (and present none of your own in return), while replying with increasing specious verbal arguments. You are coming across as a great example of the high-V low-M type I (and it would seem, you) have an issue with.

  72. @Rob

    Very smart people can explain things easily to dull people. It’s in the middle it goes wrong.

  73. Rob says:
    @res

    I assume iterated means PGS-selected embtyos to gametes to embryos made from those gametes, and on and on? That would be as effective as breeding real people selected by PGS, but much faster, right? If you are looking at 5 points/generation from embryo selection, you could probably keep getting 5 points every selection cycle, until variance has been exhausted or the children finally produced cannot carry their massive heads.

    Certainly strong selection on a large, fast-reproducing population can change a trait quite a lot very fast. Cost of that is the parents would be raising their great^n grandchildren. They won’t be as closely related as parent child. You could maybe breed more of the parents into iterated kids

    [MORE]

    You get around the the low coefficient of relatedness issue doing something like take a million overall great people. Get sperm and eggs, mate them in vitro by some method. , select embryos, make sex cells, randomly breed them… Then, instesd of selling emryos, you sell sperm/eggs from these awesome “peeople” who have been iteratively selected. I’m going to call the iteratively selected embryo’s “donors.” Sure you lose half, or whatever the correlation of one parent’s IQ and the child’s, but you have a standardish product. High end donor eggs are expensive, so these would be higher PGS scores, and you don’t even have to inconvienience a college student. Plus, you can make additions to your line every however oftern, to keep up fashion.

    Acoupke wants a baby, but dadwants a kid that’s his and so does nom. So dad fertilizes donor egg(s) and mom’s egg(s) likewise fertilized by a different donor’s sperm. Then those embryos are selected. call dad’s embryos group A and mom’s B. Gamete’s derived from group A and B are mixed with eachother to get embryos that are .25 mom and .25 dad. That’s not shabby. Plus, maybe they end up selecting embryos that have more donor genes that theirs, because they have better genetic “scores” with hopefully include more than common SNPs.

    For gay men, they both get eggs, to embryos(.5 ibd each) to gametes to final embryos(0.25) ibd. The gay men are then both grandsthers of their kid. That’s rather sweet. I that’ll be the firstbapplication of embryo to gamete techniques on people.

    With single moms, you just do donor sperm, and kid is half her. Oh crap! That gives single moms a selective advantage.

    Have you seen this post that was pulled from human varieties, but is in arvhive.is Coincidence</a

    He has charts of average PGS for some races, religions, and the average PGS of a group correlates more strongly with thr groups’ average IQs, because in big samples noise averages out. I’m interested groups far off the regression line. In the third graph, IQ vs PGS (yes, he put PGS on the y axis), but mormons have a high PGS score for their actual IQ, why? Budhists. Actually,not that interesting. Did not notice that IQ was like 10 points.

    The fourth chart, blacks have higher IQ than their PGS predicts. Does this imply blacks have some +IQ alleles that are not present in other populations, they just have theirs plus ours at lower frequency? Is this a linkagevartifact? If it is, it does not increase my faith in GWAS. If they aren’t finding causal alleles, why are the actual causal alleles not SNPs themselves. As a cartoon example, the SNP rs01 is a determined by a GWAS to be a +1 IQ allele. But it is not causal, there are a bunch of different alleles in different people close to it on the chromosome, some add, some subtract, and the weighted sum in the tested population just happens to +1? These actually causal alleles are individually rare, else they would be tested for, just like the common SNP? That would explain the tiny pluses and minuses, their just measuring some noise? Random chance about whether the rare alleles are + or minus? I would think most rare alleles would be negative. Do effect sizes for SNPs tend to go up or down with larger numbers tested? If they go down, that sounds more like noise.

    I know twin pair GWAS tend tend find smaller effect sizes than general population studies. That’s not consistent with rare variants, as they would tend to inherit the same rare variants? But the’d be even more unlinked in unrelated people.

    Most of the genome is junk, so there should be a whatever number of variants according to geometric mean of the breeding pop. Do GWAS test in true junk, like old retroviruses, or do they limit to around genes? Must be genetic regions.

    That Kirkegaard paper, i remembered wrongly. Jewish people had higher PGS (1.37σ) compared to their measured IQ (110, 0.66σ) that’s super-strange. And the Catholics and Lutherans are so close to each other it’s hard to get a feel for how IQ and PGS should measure up. Jews are not on the Councidence charts I linked to. I would think that if SNP hits were not themselves causal, but we’re linked to (rare, else they’d be SNPs) causal variants let’s call SNP s and causal variant c. When s and c are linked, they are not always inheritted together. Sometimes there’s cross over. When that happens, it is causal variant c that will increase in frequency, and that copy of s will just drift. So when a causal variant is linked to a SNP in one population, it’ll contribute to PGS equation calculated from GWAS. in the other population, with S and c unlinked, if s drifted down and c was selected, then they’d have a slightly lower PGS than their actual phenotype. So all these plus SNP just happened to unlink in and drift higher in Jews?

    Am I thinking about an artifact, and 0.66/1.37 = 0.48, is a good correlation for a PGS and group IQ. Am I doing that right? The human varieties guy got better, but he did not include Jews.

    Is this an artifact of measuring things Jews are relatively bad at, like spatial stuff? 1/3 spatial is not so excessive. I remember when verbal and maybe math IQ averaged nearly 125-130. Have American Jews been hit by serious dysgenic? Not super-dumb Jews having tons of kids, but at the high school end, marrying Gentiles so their kids regress to a lower mean, smart women fighting the patriarchy by not participating in motherhood, small families at the high end? Or smart Jews just atomizing and saying their atheists. I said smart Jews marrying Gentiles, but the same would apply to more average (for Jews) marrying Gentiles.

    Sorry for being scatterbrained today. Got a bad night’s sleep. I checked my bookmarks for blogs, and there is not much. Blogging is dead. Maybe there are substacks done by pros? I can’t pay \$5/month for every blog I like. They are going to have to work out bundles or deals, or something.

    I rather like https://halfassed.science.blog/ though he does not do PGS, he does hbd. Posts rarely, but was my source for Uriah(@crimakid on Twitter)
    Slime mold time mold is very good, especially if you want a great rundown on the obesity epidemic, which may be her(?) thing,

    • Replies: @res
  74. res says:
    @Rob

    I assume iterated means PGS-selected embtyos to gametes to embryos made from those gametes, and on and on?

    Exactly.

    Cost of that is the parents would be raising their great^n grandchildren. They won’t be as closely related as parent child.

    That’s only sort of true (big DNA blocks will be broken up, but all of the offspring will still average 50/50 mother/father). I think the bigger problem is the inbreeding (odd that Gwern does not seem to mention that), but as long as you are doing WGS (whole genome) you could probably deal with that in the screening phase.

    Have you seen this post that was pulled from human varieties, but is in arvhive.is Coincidence</a
    https://archive.is/2021.06.03-063940/https://humanvarieties.org/2021/06/01/coincidence/

    I had not. Thanks! I don’t read enough other sites (tend to rely on the good stuff finding its way here).

    The fourth chart, blacks have higher IQ than their PGS predicts. Does this imply blacks have some +IQ alleles that are not present in other populations, they just have theirs plus ours at lower frequency?

    Looking at that chart my guess is they are missing +IQ variants for Chinese. Notice how the three rightmost points have almost the same mean PGS. Then think about how the regression line would change if you raised them in line with the IQ differences.

    Any idea why that post was disappeared? We had some extensive conversations on that Piffer work here. You can search my comments if interested.

    You make some other good points there, but too many to engage with right now.

    Jewish people had higher PGS (1.37σ) compared to their measured IQ (110, 0.66σ) that’s super-strange.

    Agreed. The other strange thing is the Jewish IQ distribution in Figure 1. Two humps about an SD apart and a fat left tail.

    I rather like https://halfassed.science.blog/ though he does not do PGS, he does hbd. Posts rarely, but was my source for Uriah(@crimakid on Twitter)
    Slime mold time mold is very good, especially if you want a great rundown on the obesity epidemic, which may be her(?) thing,

    Thanks. Some good stuff at the first link. This alone was worth the trouble of skimming.
    https://halfassed.science.blog/2021/01/30/what-explains-the-south-north-gradient-in-german-iq/

    In Germany however, there is a distinctive IQ gradient from South to North, while for height it is the other way around. Because of the Alpes in the South and the Sea in the North, the South has a harsher climate, which would serve to explain the IQ gradient. But why does the height gradient not follow suite?

    Probably, because selection for IQ is much more recent than selection for height. Height difference within Europe are thousands of years old, as mentioned in Roman texts. IQ differences are likely significantly more recent.

    But I doubt that IQ differences that established themselves in the last 1000 years within Europe owe much to climate. Instead I have a different theory: I propose that iodine deficiency had a strong positive selection effect on IQ.

    Slime mold time mold also looks interesting, but I was unable to resolve your Uriah reference (e.g. don’t see crimakid on Twitter, did he get banned?).

    Thanks!

    • Replies: @Rob
  75. Factorize says:
    @res

    gwern mentions that large phenotype uplifts could occur if gamete (especially spermatids) could be non-destructively genotyped. If one had access to essentially an unlimited number of such sperm one would then have the potential to optimize a haploid of a human’s genome. Yet, this might be already possible. If one were able to genotype 3 of 4 connected spermatids during telophase II of meiosis II, then one would be able to infer the exact genotype non-destructively of the other spermatid. If the genomes of one million (on average) of such spermatids could be inferred, then one would roughly have the optimized haploid genome of a man. A similar strategy could apply with the polar bodies of females, though owing to the difficulty in obtaining large numbers of egg cells, such an optimization might only allow for finding the best egg cell of 10 instead of the optimal of essentially an infinity of sperm for males. The chosen optimized sperm could fertilize the best egg cell through ICSI and this might allow for substantial and rapid phenotype selection and enhancement.

    During each succeeding generation optimal sperm selection could occur and the sperm could then act as a genome vector for the optimized chromosomes to the female lines. So, in the first generation the daughter (and the sons) would receive 23 selected chromosomes. When this first generation mated, the males with another contribution of an optimized haploid genome would transfer these 23 selected chromosomes to the embryos that would become female (perhaps ~12 on average without egg selection and with 10:1 selection possibly 15 of the selected chromosomes from the mothers would also be inherited). The females of the second generation would have possibly 38 of 46 selected chromosomes. Within even ~3 generations a strategy with optimized sperm selection would create humans with fully selected genomes. The sperm from the males is adding 100% selected chromosomes (i.e., 23 selected chromosomes) every generation! Selected chromosomes would rapidly displace the non-selected chromosomes.

    This might be the most rapid method to optimize human genetics with seemingly low end reproductive technology: the only unknown here is whether the 3 of 4 spermatids could be identified and sequenced while retaining the fourth spermatid with inferred genotype. With a large enough number of sperm, the fitness difference between the number 1 and 2 rated sperm (of perhaps 1,000) probably would not be that large. Perhaps even the difference in fitness of the egg cells might then be considered less important as after the first generation or two all the chromosomes would be selected for both parents. Here the heavy lifting of selection is done mostly through sperm.

    In such a scenario, the reproductive outcome is less constrained by embryo choice. In fact, using this gamete selection idea the entire concern about “embryo abortion” might be overcome. If you can choose gametes, then there is then no obvious need to have to choose from a large number of embryos anymore. You could simply start with the fittest sperm and fittest egg perform ICSI with the intention of implanting this single embryo. There would no longer be any rational of having a whole rack of embryos to select among. The economics involved would actually create a large incentive to use a single embryo, if possible.

  76. Rob says:
    @res

    Wait, Gwern’s iterated embryo selection is wannabe mom and date produce some gametes. Those gametes are combined into, say 10 embryos. Those embryos produce/turn into gametes, and those gametes are combined with each other? So they are hyperinbred? I just assumed it was SIMS for selection. That a bunch of people’s embryo-derived gametes formed a “breeding population” with generations on the order of weeks or months, not 20 years.

    I am not sure super-inbreeding in vitro would work out. I have heard that the average person is a carrier for like 2-7 lethal recessives. But that’s just lethal. Look at Saudis or Pakistanis in Britain to see the effects of non-lethal negative alleles. And they are not siblings. In vitro, only alleles that kill cells generally, embryonic stem cells specifically or sex cells and whatever stages are between the last two. Neural defects would not be weeded out, for one.

    [MORE]

    Compare that to Aleithea Inc buying gametes from their exclusively contracted group of athletes scientists, artists, actors, brilliant engineers. Hell, just everyone who applies with an IQ above, say, 145. If men were rational animals, then they would pay to give samples. Probably women too. From women, turn skin or deeper cells embryonic, then turn to ova. No egg donation regime necessary. Then these thousand human donors’ gametes are randomly mixed (to make the math easier) not in one big tank, but randomly determined, and keep track. Make however many embryos from each human donor pair, score those embryos. Keep some from each, according to score, which might vary for different pairings: don’t look for high IQ in the actor-pole vaulter embryos. Don’t look for beauty and coordination in the scientist-engineer pairs. Then do panmixia again, except avoid close inbreeding. You could probably repeat the process maybe 10 generations and have a very low percentage that is identical by descent. So 10 generations of maybe 0.5 σ each iteration on a couple/three phenotypes, IQ, athleticism… so 5 σ on 3 fairly complex phenotypes. Here’s why you don’t care about inbreeding – you are not selling embryos, you are selling gametes! Inbreeding the fetuses doesn’t damage the germ cells. Inbreeding goes away with one outbreeding. Not to mention, let’s say your RIMS1 IQ mod, cuz genetic engineering and iterated embryo selection could be combined, is only beneficial in heterozygotes, actually, you could just engineer multiple chromosomes with it, but getting homogeneous germ cells from embryonic tissue when homozygosity would kill a person is quality control. Plus, Aleithea Inc probably likes customers’ kids having to come to them to reproduce safely.

    Aleithea takes their favorite mixes of traits and makes a catalog of gametes or “parental fetuses, even uses the best in silico simulation to make Jennifer real. And she is beautiful, genotypic IQ, no tendency to fat or hair loss, no lethal recessives. If she’s not inbred, you can even grow Jennifer into a real person in a surrogate. You use her as your spokesperson. Follow her through life, as a cute baby who does not cry much, who sleeps through the night. Her years as a curious toddler, where she’s affectionate and obedient to her parents, with just the right amount of independence. Her tween years, her teen years, her acceptance to Harvard…The eggs of her clones are available for just 50,000 dollars.

    Then you buy Jennifer’s egg your wife buys Rico’s sperm, you get an embryo, she gets an embryo, you make gametes from them you then in nine months you have a baby.

    You know, if they can derive gametes from embryonic stem cells, then there should not be any reason you could not make eggs from male cells. I doubt the extra X is required to create an egg. Sure there would be Y-bearing eggs, but if they’re fertilized by a Y-sperm, the cell will die. If a Y egg is fertilized by an X sperm you get a boy, a reverse boy. A literal trans boy. So a gay male couple could have their own kid who is 50% each of them. There are too many genes necessary for female embryonic stem cells to be turned into sperm. Maybe enough could be delivered as mRNA?

    I’m sorry Uriah is https://twitter.com/crimkadid
    He probably won’t be banned soon, as very few people follow him. He’s no BAP, who I didn’t much care for.
    At least in the past, Uriah did not thread his essays, so if you scroll his feed they read backward. https://threadreaderapp.com/user/crimkadid has what I assume is the greatest hits.

    Did you ever read Counting Heads and/or Mind Over Ship by David Marusek? He has a plausible-feeling future incorporating mostly realistic human engineering. Book 1 is set on earth, book two has some action on a torus in solar orbit.

    If you’re a bit of a dork, GURPS Transhuman Space is a somewhat plausible setting with absolutely no magic, woo, or psi. There’s some exploration of what genetically engineering people would do, but it is way conservative, considering a 25 point IQ increase is plausibly 15-20 years away.

    I’ll go thru my big collection of papers and ebooks I have mostly not read and see if I can find anything I think might appeal to you.

    Professor Thompson, thank you for letting us use your thread so off-topically.

Current Commenter
says:

Leave a Reply -


 Remember My InformationWhy?
 Email Replies to my Comment
$
Submitted comments have been licensed to The Unz Review and may be republished elsewhere at the sole discretion of the latter
Commenting Disabled While in Translation Mode
Subscribe to This Comment Thread via RSS Subscribe to All James Thompson Comments via RSS