What percentage of the b-w gap are PGS able to predict? How do we account for differing r2 for blacks and whites on pgs_edu when predicting the iq gap mediated by allele variants
Your estimates of heritability are hugely exaggerated. While behavior geneticists believed that IQ was 80% heritable in the 90s, more recent research has shown that the heritability has to be much lower. IQ scores are rising all the time, a fact that isn't compatible with an 80% heritability. Adoption studies have shown massive IQ gains for children. New theories of environmental influence on IQ have become increasingly accepted. Read about it here: https://open.substack.com/pub/eclecticinquiries/p/on-race-racism-iq-and-heritability?r=4952v2&utm_campaign=post&utm_medium=web
There were too many comments on your twitter thread to read. A number of people pointed this out and I agree. The IQ gap between black and white Americans is likely entirely genetic. There are no starving people in the US nor do we have lead paint or any widespread pollutants.
It's like sex differences. The more egalitarian a society becomes, the more of the differences we see are genetic. Women can be anything in the US or Western Europe. Yet so few of them go into computer science or engineering. Instead a poor country like Romania has a lot of women engineers.
But the IQ gap between Sub Saharan Africans in Africa and whites probably has a sizable environmental explanation. Giving the Raven's test to children after they spend a day mining cobalt is not exact good data right? There's still a lot of pollution in their drinking water also.
Bundling IQ scores together with hard data on polygenic score clusters in populations in order to support a hereditarian claim on intelligence is like constructing a CDO from a combination of entirely different tranches and having it all certified AA. Kidney functioning is straightforward, and the parameters used to verify its range of normal functioning are well-defined, as is the balance of interactions. The functioning can be mapped as an organic process, in all of its relevant aspects. Human intelligence is a lot more elusive than kidney function. Even its definition is elusive.
As a rule, IQ tests take for granted that "abstract reasoning" is the substrate of intelligence. I contend that the aspect of intelligence measured by Raven's Progressive Matrices is per se no more probative as a linear measure of general intelligence than the ability to compose an especially complex musical passage is per se indicative of general musical ability. It's an aspect. It is not the sine qua non of superior musical sensitivity, or creativity, or a wider facility with the art form. Don't get me wrong, I respect abstract reasoning ability, and agree that some quotient of it is required as an interaction with other features of intelligence. It's definitely an asset, and superior ability is crucial for success in some occupational fields. But just because someone is able to exhibit superior performance in abstract reasoning, that does not automatically make them the smartest person in the room. And if the person has deficits in other aspects of intelligence, that ability alone is of a narrowly delimited value.
Beyond that, a score on an IQ test is simply not a direct, first order biological measure in the same sense as a creatinine level and the wider implications of its effect on kidney function. Income and educational attainment is not that sort of metric, either. Yet it's often taken for granted as providing similar validity! That assumption deserves a lot more examination than it's gotten, because it's bogus. Just because both creatinine tests and IQ tests express their results as number scores, that does not lead to the conclusion that those numbers have equal probative value. But that's the assumption undergirding the matching of IQ scores of discrete subpopulations and correlating them with polygenic gene cluster frequencies of subpopulations that are abstracted from first order measurements taken from individual humans. That calculation doesn't have much in common with the scientific and medical study of biologic functioning, or with genetics. It's quite an inferential leap, from aggregating polygenic scores to find indentifiable clusters by population ancestry to the more ambitious inquiry into how they might happen to correlate to traits like "performance on IQ tests", "income", and "educational attainment." The first part of that attempted linkage is straightforward genetics; the second is an extrapolation based on other extrapolations. Less surely grounded data, in terms of its pliability of evaluation and the possibility of confounding factors. Two different tranches, bundled together despite their dissimilarities, using methodologies that might arguably be viewed as constructed more for their convenience for the researchers than their ability to provide precise answers and identify the biological mechanisms at work in individuals.
That sort of research exhibits quite a different character from, say, genetic markers related to biological variations in kidney function, or aerobic capacity. Which relates to a couple of other problematic features of intelligence evaluation research: is there a place for groupthink when it comes to evaluating the intelligence of an individual? Can the complete spectrum and potential of mental ability of all individual humans really be precisely and reliably scored on the basis of their record of performance on a handful of formal examinations of their abstract reasoning and their command of scholastic skill sets? Even the testing industry agrees that dynamic factors like lack of sleep and lack of motivation can negatively impact performance on these tests by the same individual.
It's worth noting how these researches couch their own studies: in terms of leads and indications, not sure conclusions. Again, not with the same level of knowledge assurance about the the vital operating mechanisms and parameters of biological functions that have been mapped by medical science, and therefore readily amenable to assessment with chemical analysis in a laboratory. There's no blood test for intelligence. Answers to timed exam questions are not the same thing.
"Sibling regression toward the mean is one line: a black sibling of someone with an IQ of 120 will have an IQ of about 100, while a white sibling would have an IQ of about 110 (Jensen 1973, 118). "
Interesting. Is there also data about the IQ of the parents of this individual with 120 IQ?
What is the distribution of individuals with 120+ IQ in the black community?
If the individuals are a super minority, then the genetic impact might be less pronounced, as scarcity might limit reproduction partners.
If both parents have an IQ above 120, I would be curious to see if this sibling regression relationships remain the same.
I know that some level of endogamy exists between social classes, which leads to IQ wealth being concentrated within communities. Elites marry other elites, and high IQ individuals rarely interact outside similar groups. I wonder if the observed sibling regression is based on the expectation that this social endogamy is the same across both communities.
I also question the actual accuracy of income as a proxy for measuring sibling regression within the black community - DEI and affirmative action type interventions must have an outsize effect on the output on that analysis. I would be curious to know how one would correct for such interventions.
Really though, I'm curious to see what newer data looks like. A lot of the African Elite professionals have migrated to the west over the last twenty years, I expect changes, at least to our understanding of heritability, race and environment.
I understand why people can have concerns over this question. But it is possible, in a free society, that defends civil rights, to have this discussion safely, without threatening anyone. We all live gifted, and burdened with our innate traits. There are better and worse traits. We all know that about ourselves, our loved ones, and everyone else. And we see the gifts and lesions of each other. I think most of us do appreciate the diversity that blesses and plagues each of us. But a few traits separate us- the honesty and courage to admit these things.
"For example, the black-white IQ gap remains roughly the same at all income levels (Herrnstein & Murray 1994, 288), meaning lower group income couldn’t be the cause."
Doesn't that mean that although IQ predicts income (to some degree), Blacks on average, require a lower IQ score than do whites to achieve the same level of income? And wouldn't that then mean that the test actually IS biased -- such that a lower score for African Americans ultimately has the same consequences for life success as a higher score for whites? Am I wrong?
and, second:
"When they have the same IQ as whites, blacks are more likely than whites to graduate from college and to attain a high status occupation. "
Again (same basic argument as above regarding wealth). Doesn't this finding mean that African Americans with a lower IQ score are as academically able as whites with a somewhat higher IQ score? That IS what the sentence I quoted means, isn't it? If not -- then what DOES it mean?? And if I've interpreted the finding correctly, then wouldn't that mean there is a degree of bias in the test, such that lower IQ scores for African Americans are predictive of the same degree of academic and life success as high scores for whites? What am I missing here?
"Another line of evidence is that the blacks have a smaller standard deviation for IQ than whites (Jensen 1973, 212)."
Don't think so. Your own table from our study above shows slightly larger Black SD.
What percentage of the b-w gap are PGS able to predict? How do we account for differing r2 for blacks and whites on pgs_edu when predicting the iq gap mediated by allele variants
https://russellwarne.com/2022/12/17/irish-iq-the-massive-rise-that-never-happened/
the irish iq has always been relatively high (98)
Your estimates of heritability are hugely exaggerated. While behavior geneticists believed that IQ was 80% heritable in the 90s, more recent research has shown that the heritability has to be much lower. IQ scores are rising all the time, a fact that isn't compatible with an 80% heritability. Adoption studies have shown massive IQ gains for children. New theories of environmental influence on IQ have become increasingly accepted. Read about it here: https://open.substack.com/pub/eclecticinquiries/p/on-race-racism-iq-and-heritability?r=4952v2&utm_campaign=post&utm_medium=web
Alright, go live in Africa with only West Africans and don't come back.
There were too many comments on your twitter thread to read. A number of people pointed this out and I agree. The IQ gap between black and white Americans is likely entirely genetic. There are no starving people in the US nor do we have lead paint or any widespread pollutants.
It's like sex differences. The more egalitarian a society becomes, the more of the differences we see are genetic. Women can be anything in the US or Western Europe. Yet so few of them go into computer science or engineering. Instead a poor country like Romania has a lot of women engineers.
But the IQ gap between Sub Saharan Africans in Africa and whites probably has a sizable environmental explanation. Giving the Raven's test to children after they spend a day mining cobalt is not exact good data right? There's still a lot of pollution in their drinking water also.
banned for llm spam
Bundling IQ scores together with hard data on polygenic score clusters in populations in order to support a hereditarian claim on intelligence is like constructing a CDO from a combination of entirely different tranches and having it all certified AA. Kidney functioning is straightforward, and the parameters used to verify its range of normal functioning are well-defined, as is the balance of interactions. The functioning can be mapped as an organic process, in all of its relevant aspects. Human intelligence is a lot more elusive than kidney function. Even its definition is elusive.
As a rule, IQ tests take for granted that "abstract reasoning" is the substrate of intelligence. I contend that the aspect of intelligence measured by Raven's Progressive Matrices is per se no more probative as a linear measure of general intelligence than the ability to compose an especially complex musical passage is per se indicative of general musical ability. It's an aspect. It is not the sine qua non of superior musical sensitivity, or creativity, or a wider facility with the art form. Don't get me wrong, I respect abstract reasoning ability, and agree that some quotient of it is required as an interaction with other features of intelligence. It's definitely an asset, and superior ability is crucial for success in some occupational fields. But just because someone is able to exhibit superior performance in abstract reasoning, that does not automatically make them the smartest person in the room. And if the person has deficits in other aspects of intelligence, that ability alone is of a narrowly delimited value.
Beyond that, a score on an IQ test is simply not a direct, first order biological measure in the same sense as a creatinine level and the wider implications of its effect on kidney function. Income and educational attainment is not that sort of metric, either. Yet it's often taken for granted as providing similar validity! That assumption deserves a lot more examination than it's gotten, because it's bogus. Just because both creatinine tests and IQ tests express their results as number scores, that does not lead to the conclusion that those numbers have equal probative value. But that's the assumption undergirding the matching of IQ scores of discrete subpopulations and correlating them with polygenic gene cluster frequencies of subpopulations that are abstracted from first order measurements taken from individual humans. That calculation doesn't have much in common with the scientific and medical study of biologic functioning, or with genetics. It's quite an inferential leap, from aggregating polygenic scores to find indentifiable clusters by population ancestry to the more ambitious inquiry into how they might happen to correlate to traits like "performance on IQ tests", "income", and "educational attainment." The first part of that attempted linkage is straightforward genetics; the second is an extrapolation based on other extrapolations. Less surely grounded data, in terms of its pliability of evaluation and the possibility of confounding factors. Two different tranches, bundled together despite their dissimilarities, using methodologies that might arguably be viewed as constructed more for their convenience for the researchers than their ability to provide precise answers and identify the biological mechanisms at work in individuals.
That sort of research exhibits quite a different character from, say, genetic markers related to biological variations in kidney function, or aerobic capacity. Which relates to a couple of other problematic features of intelligence evaluation research: is there a place for groupthink when it comes to evaluating the intelligence of an individual? Can the complete spectrum and potential of mental ability of all individual humans really be precisely and reliably scored on the basis of their record of performance on a handful of formal examinations of their abstract reasoning and their command of scholastic skill sets? Even the testing industry agrees that dynamic factors like lack of sleep and lack of motivation can negatively impact performance on these tests by the same individual.
(2022) https://onlinelibrary.wiley.com/doi/full/10.1111/jsr.13815
(2011) https://www.nih.gov/news-events/nih-research-matters/motivation-may-influence-iq-scores
It's worth noting how these researches couch their own studies: in terms of leads and indications, not sure conclusions. Again, not with the same level of knowledge assurance about the the vital operating mechanisms and parameters of biological functions that have been mapped by medical science, and therefore readily amenable to assessment with chemical analysis in a laboratory. There's no blood test for intelligence. Answers to timed exam questions are not the same thing.
continued here https://theinfinitesimal.substack.com/p/no-intelligence-is-not-like-height/comment/69621729?utm_source=activity_item
"Sibling regression toward the mean is one line: a black sibling of someone with an IQ of 120 will have an IQ of about 100, while a white sibling would have an IQ of about 110 (Jensen 1973, 118). "
Interesting. Is there also data about the IQ of the parents of this individual with 120 IQ?
What is the distribution of individuals with 120+ IQ in the black community?
If the individuals are a super minority, then the genetic impact might be less pronounced, as scarcity might limit reproduction partners.
If both parents have an IQ above 120, I would be curious to see if this sibling regression relationships remain the same.
I know that some level of endogamy exists between social classes, which leads to IQ wealth being concentrated within communities. Elites marry other elites, and high IQ individuals rarely interact outside similar groups. I wonder if the observed sibling regression is based on the expectation that this social endogamy is the same across both communities.
I also question the actual accuracy of income as a proxy for measuring sibling regression within the black community - DEI and affirmative action type interventions must have an outsize effect on the output on that analysis. I would be curious to know how one would correct for such interventions.
Really though, I'm curious to see what newer data looks like. A lot of the African Elite professionals have migrated to the west over the last twenty years, I expect changes, at least to our understanding of heritability, race and environment.
I understand why people can have concerns over this question. But it is possible, in a free society, that defends civil rights, to have this discussion safely, without threatening anyone. We all live gifted, and burdened with our innate traits. There are better and worse traits. We all know that about ourselves, our loved ones, and everyone else. And we see the gifts and lesions of each other. I think most of us do appreciate the diversity that blesses and plagues each of us. But a few traits separate us- the honesty and courage to admit these things.
Two of the findings you reference puzzle me.
First:
"For example, the black-white IQ gap remains roughly the same at all income levels (Herrnstein & Murray 1994, 288), meaning lower group income couldn’t be the cause."
Doesn't that mean that although IQ predicts income (to some degree), Blacks on average, require a lower IQ score than do whites to achieve the same level of income? And wouldn't that then mean that the test actually IS biased -- such that a lower score for African Americans ultimately has the same consequences for life success as a higher score for whites? Am I wrong?
and, second:
"When they have the same IQ as whites, blacks are more likely than whites to graduate from college and to attain a high status occupation. "
Again (same basic argument as above regarding wealth). Doesn't this finding mean that African Americans with a lower IQ score are as academically able as whites with a somewhat higher IQ score? That IS what the sentence I quoted means, isn't it? If not -- then what DOES it mean?? And if I've interpreted the finding correctly, then wouldn't that mean there is a degree of bias in the test, such that lower IQ scores for African Americans are predictive of the same degree of academic and life success as high scores for whites? What am I missing here?
You misunderstand. One association is one own's income vs IQ (when adult), other is parental income vs child IQ.