Unfortunately, the research I have seen disagrees. Race proxies for quite a bit more than economic status- for example, approaches to education vary significantly by culture, and race and culture track pretty well. Beyond that, if there are genetic effects we would expect to see them more strongly in race comparisons than economic status comparisons.
Beyond that, if there are genetic effects we would expect to see them more strongly in race comparisons than economic status comparisons.
I strongly disagree. The relationship between economic status and intelligence is maintained by strong feedback effects. Intelligent people tend to move to higher economic status, and then mostly pass on those genes for intelligence to offspring. But then, if the offspring were unlucky enough to not get the genes, then they will tend to slide out of that economic status class.
The genetic aspect of the relationship between race and intelligence, if it exists at all, is something of a historical accident. There may be a degree of stability to the relationship, but there is no active feedback loop maintaining it. People do not slide from one racial grouping to another simply because they are born with more or less than the expected amount of intelligence.
Notice that I am not disputing your research claim regarding the degree to which race and economics serve as predictors of measured intelligence. I only dispute the validity of your intuitions regarding what “we would expect to see”.
You put together a strong argument. I think that this will depend on which genes we examine; if there is some intelligence-boosting gene that most Xs have and most Ys do not, then the effect of that gene will be more noticeable when we do race comparisons than economic status comparisons. However, intragroup variability is higher than intergroup variability, which suggests that for every intelligence-boosting gene like that there should be at least three which aren’t strongly associated with race, and if we accounted for all the genes then the intelligence-economic status connection would likely be stronger than the intelligence-race connection (though I have insufficient data to conclude it would be stronger).
I was thinking about the first kind of genetic effects only, since they seem easier to find and study than the second, but you are right to question that choice.
I think we are in rough agreement, though I think you stated this backwards:
However, intergroup variability is higher than intragroup variability
Actually, and empirically, variability within a group is relatively high and variation between groups rather low. (More technically, group membership explains only a small fraction of the total variance.) And you seem to understand this below, with your estimate that three out of four genes bearing on intelligence will not be closely correlated with race.
Thanks! I did, fixed. I should also mention that I’m being loose by assuming all genes that impact intelligence do so by roughly the same amount, which is obviously not true.
We may need to be more explicit about the claim under discussion. I intended to say that if you partitioned students by race and economic status, ran a regression, then also added in some gene markers and ran another regression, you would find the race coefficient decreased more than the economic status coefficient.
The data you are providing suggests that genes become a better way to differentiate students at higher economic status from each other, which does not appear to disagree with my claim.
I will also note that the mechanism behind your data- increased uniformity of parenting styles- might also be strongly noticeable when looking at race instead of economic status.
We may need to be more explicit about the claim under discussion. I intended to say that if you partitioned students by race and economic status, ran a regression, then also added in some gene markers and ran another regression, you would find the race coefficient decreased more than the economic status coefficient.
Oh. That makes sense then, though there’s the question of whether you’ve picked the relevant genetic markers.
I have always seen socioeconomic status as the best predictor of a student’s success in education. You do have a point about differing cultural approaches, but after a generation or two, it seems like culture and socioeconomic status would correlate. Success in education improves socioeconomic status, which further improves success in education for the next generation, even if cultural norms drove the parents’ success.
I haven’t seen any compelling evidence for genetic effects; “race” is much more of a social construction than a genetic one. There are some genetic markers and traits that vary more between races than individuals, but overall genetic variation within a specific population is much greater than the variation between races.
There are some genetic markers and traits that vary more between races than individuals, but overall genetic variation within a specific population is much greater than the variation between races.
But it doesn’t follow that race couldn’t possibly be a useful biological classification. Even if the distributions of two groups overlap strongly when any one trait or marker is considered, the groups might still be easily distinguished when you look at the whole configuration space—consider an illustrative diagram.
You do have a point about differing cultural approaches, but after a generation or two, it seems like culture and socioeconomic status would correlate.
Right, but they would need to be identical, not just correlate, for a model that only includes one to be as good as a model that includes both.
Accounting for race is generally just a lazy way to account for economic status.
Unfortunately, the research I have seen disagrees. Race proxies for quite a bit more than economic status- for example, approaches to education vary significantly by culture, and race and culture track pretty well. Beyond that, if there are genetic effects we would expect to see them more strongly in race comparisons than economic status comparisons.
I strongly disagree. The relationship between economic status and intelligence is maintained by strong feedback effects. Intelligent people tend to move to higher economic status, and then mostly pass on those genes for intelligence to offspring. But then, if the offspring were unlucky enough to not get the genes, then they will tend to slide out of that economic status class.
The genetic aspect of the relationship between race and intelligence, if it exists at all, is something of a historical accident. There may be a degree of stability to the relationship, but there is no active feedback loop maintaining it. People do not slide from one racial grouping to another simply because they are born with more or less than the expected amount of intelligence.
Notice that I am not disputing your research claim regarding the degree to which race and economics serve as predictors of measured intelligence. I only dispute the validity of your intuitions regarding what “we would expect to see”.
You put together a strong argument. I think that this will depend on which genes we examine; if there is some intelligence-boosting gene that most Xs have and most Ys do not, then the effect of that gene will be more noticeable when we do race comparisons than economic status comparisons. However, intragroup variability is higher than intergroup variability, which suggests that for every intelligence-boosting gene like that there should be at least three which aren’t strongly associated with race, and if we accounted for all the genes then the intelligence-economic status connection would likely be stronger than the intelligence-race connection (though I have insufficient data to conclude it would be stronger).
I was thinking about the first kind of genetic effects only, since they seem easier to find and study than the second, but you are right to question that choice.
I think we are in rough agreement, though I think you stated this backwards:
Actually, and empirically, variability within a group is relatively high and variation between groups rather low. (More technically, group membership explains only a small fraction of the total variance.) And you seem to understand this below, with your estimate that three out of four genes bearing on intelligence will not be closely correlated with race.
Thanks! I did, fixed. I should also mention that I’m being loose by assuming all genes that impact intelligence do so by roughly the same amount, which is obviously not true.
I would expect the opposite to be true, based on results that genetic effects are most important at high economic status.
We may need to be more explicit about the claim under discussion. I intended to say that if you partitioned students by race and economic status, ran a regression, then also added in some gene markers and ran another regression, you would find the race coefficient decreased more than the economic status coefficient.
The data you are providing suggests that genes become a better way to differentiate students at higher economic status from each other, which does not appear to disagree with my claim.
I will also note that the mechanism behind your data- increased uniformity of parenting styles- might also be strongly noticeable when looking at race instead of economic status.
Oh. That makes sense then, though there’s the question of whether you’ve picked the relevant genetic markers.
I have always seen socioeconomic status as the best predictor of a student’s success in education. You do have a point about differing cultural approaches, but after a generation or two, it seems like culture and socioeconomic status would correlate. Success in education improves socioeconomic status, which further improves success in education for the next generation, even if cultural norms drove the parents’ success.
I haven’t seen any compelling evidence for genetic effects; “race” is much more of a social construction than a genetic one. There are some genetic markers and traits that vary more between races than individuals, but overall genetic variation within a specific population is much greater than the variation between races.
But it doesn’t follow that race couldn’t possibly be a useful biological classification. Even if the distributions of two groups overlap strongly when any one trait or marker is considered, the groups might still be easily distinguished when you look at the whole configuration space—consider an illustrative diagram.
Right, but they would need to be identical, not just correlate, for a model that only includes one to be as good as a model that includes both.