Race & IQ

Dan MacArthur's post Should scientists study race and IQ? points to a debate in Nature. Well, scientists are studying genetic variation. And others are engaged in the project of psychometrics. This seems to fall into the category of the economist predicting the past perfectly. Lagging indicators & all. I mean, if we're talking about whole genome sequencing of newborns in 2019....

Tags

More like this

No since they can't find a way of assessing intelligence that is free from cultural or gender biases so that racist and/or bigoted cranks won't use the data to push their agenda. The "IQ" fails for being subject to such biases, not to mention being an impermanent measure of someone's intellectual capacity.

The "IQ" fails for being subject to such biases, not to mention being an impermanent measure of someone's intellectual capacity.

not really. most 'biases' are also adduced in a circular manner.

No you can demonstrate that one's IQ changes with time, with educational attainment, with training taking tests etc. This has been known since Binet invented the test and he opposed the American interpretation of the IQ as some kind of fixed quantity. The test was made to identify those who needed help, not to affix a stamp on people that says "stupid". Americans have been abusing it for decades.

If you compare races without very carefully controlling for these facts that there are huge socioeconomic differences (and I think ultimately we can never really account for all the cultural inputs) all you're doing is studying poverty and culture. Every goddamn time.

and I think ultimately we can never really account for all the cultural inputs

sure. some cultures probably produce stupidity. and i am aware of the flynn effect. norm of reaction doesn't negate that IQ is around half heritable. you can call it something else. i.e., time-it-takes-to-do-an-algebraic-manipulation. doesn't really matter.

"No since they can't find a way of assessing intelligence that is free from cultural or gender biases so that racist and/or bigoted cranks won't use the data to push their agenda."

I dunno. There are groups which have been hideously discriminated against historically who nonetheless manage to blow your average WASP out of the water in terms of average IQ. And if raw bias was the issue it should be possible to use iterative methods to separate out test questions that had an unusually large race/class/gender/whatever gap in the answers. So far as I know, no one has yet managed to do this and create a test that doesn't show certain group differences despite considerable effort.

And the next paragraph:

The effect of restriction of range on IQ was examined by Matt McGue and colleagues, who wrote that "restriction in range in parent disinhibitory psychopathology and family SES had no effect on adoptive-sibling correlations ... IQ."[10] On the other hand, a 2003 study by Eric Turkheimer, Andreana Haley, Mary Waldron, Brian D'Onofrio, Irving I. Gottesman demonstrated that the proportions of IQ variance attributable to genes and environment vary with socioeconomic status. They found that in impoverished families, 60% of the variance in IQ "in a sample of 7-year-old twins" is accounted for by the shared environment, and the contribution of genes was close to zero.[11]

one's IQ changes with time

IQ scores are extremely stable between childhood and adulthood (r = ~.8 to >.9 in various studies).

with educational attainment

childhood IQ predicts adult educational attainment, and in multivariant studies there is a strong shared genetic component to phenotypic variance in both IQ and educational attainment within populations. there are a few studies that also add population and gender differences to these multivariant models. i'll leave it as an exercise for the interested reader to look at the work of David Rowe and others in this area.

training taking tests

training reduces the correlation between performance on the particular trained task and other outcomes such as educational attainment or other IQ test items. there's no evidence from the relationship between individual and group differences in performance on particular items that the differences in average performance between groups are a function of test training. the rank order of IQ test item difficulty is extremely conserved between gender and ethnic groups.

7-year-old twins

the heritability of IQ is much lower in 7 year olds than in 18 year olds. many effects, such as early childhood interventions, fade out by adulthood. group differences in IQ are also more pronounced among adults than among children. cause of this effect is still largely a matter of speculation.

a way of assessing intelligence that is free from cultural or gender biases

there is no obvious predictive bias seen in the relationship between IQ scores and outcomes such as educational attainment among gender and ethnic groups in the US. this doesn't rule out more esoteric effects, but in the sense you seem to have meant it, there is no evidence for bias in IQ testing.

only a white male with privilege would have the gall to go by a handle like "lol" and presume that their assertions might be taken seriously.

Razib said ".. i'm well aware of the reality that heritability of IQ varies with SES."

So does paternity certainty and perhaps the level of assortative mating.

von

By vonagan cheeseman (not verified) on 17 Feb 2009 #permalink

Speaking of whites, one characteristic of the latest wave of HBD is the presence of nonwhite, specifically Asian, theorists and supporters: Bruce Lahn, Satoshi Kanazawa, Stephen Hsu...

That's another demographic distinction from earlier waves, in addition to the presence of biologists & biological anthropologists and political liberals. In other words, it has emerged from the psychometry (Jensen, Brand) and white nationalist (Rushton, MacDonald, Salter) ghettos.

Blahblahblah Gould blahblahblahblah Flynn effect blahdidiblah Hitler blahblahablah social darwinism blahblahblah algerian soldiers blahblahblah you evil racist scum.

I'm going to pull a Clinton and say that it depends on the meaning of 'should'. I don't think that is ethically or morally wrong for scientists to investigate anything. However, since both IQ and race are fairly subjective, it will be difficult to design a meaningful study, and it it's very likely that any study would only confirm the beliefs that the researchers already hold. Even a fair study would be unlikely to change anyone's mind. So scientists probably shouldn't study it because they could spend their time and money on something more productive.

catgirl et al,

So in essence you're saying scientists shouldn't study anything that might offend some group or other. By that logic, evolution, particle physics which both offend fundamentalists shouldn't be studied, and what will be next - there's always a lobby who doesn't like you to study some topic. If scientists were to listen to them, we'd still be living in caves - wake up!

Second, what could be more productive that defeating racism by exposing the true underpinnings of differential outcomes!

particle physics offends fundamentalists?

Second, what could be more productive that defeating racism by exposing the true underpinnings of differential outcomes!

Black people are stupider not b/c of their skin color but b/c of their genes! This will end racism.

By Ando Hoba (not verified) on 17 Feb 2009 #permalink

Ando Hoba,

Do you think genes are substantially involved in differential outcomes of individuals and groups?

If not, why not??

"IQ scores are extremely stable between childhood and adulthood (r = ~.8 to >.9 in various studies)."

lol,

What about the systematic increase in the B-W IQ gap from childhood to adulthood?

By Unreliable Source (not verified) on 17 Feb 2009 #permalink

There are some serious people out there whose work is worth discussing, like Turkheimer or Nisbett. Why can't the folks who are critical of the prevailing genetic slant here tell us about what these guys say instead of the silly stuff that floods discussions here.

Henry

By Henry Harpending (not verified) on 17 Feb 2009 #permalink

So was Watson correct when he said nothing will ever work in Africa because they have lower IQs? Or do you believe as I do that the application of tests across cultures is inherently flawed? This was mentioned in the "for" essay as well. To me, that's prima facie evidence that the test is highly subject to cultural biases and socioeconomic factors, and genetics plays a small effect. After all, their descendants in this country, while scoring lower than whites or asians, do a lot better. How about the Flynn effect? It's being dismissed pretty glibly here. But do you really think it's a genetic factor or are social factors different?

And again I'd point out the terrible history of these endeavors, as described in the "against" essay, and in Mismeasure. There is just no value in these studies, and the results historically are always biased. It strikes me as hubris to assume we are past these biases, and that somehow the research won't be used again to put down one race or another.

Finally, Razib, it's one thing to acknowledge the problem, it's another to address it. It keeps on being stated as fact that cultural factors have somehow been eliminated, but I see again and again studies on subpopulations showing ways in which the larger population studies are failing to account for these differences (again as mentioned in the against article).

Or do you believe as I do that the application of tests across cultures is inherently flawed?

i don't believe it is inherently flawed. though i do grant some cultures probably make people more retarded than they would otherwise be.

How about the Flynn effect? It's being dismissed pretty glibly here. But do you really think it's a genetic factor or are social factors different?

i'm not a psychologist. i could believe, as flynn does, people are getting better at these tests because of social inputs.

And again I'd point out the terrible history of these endeavors, as described in the "against" essay, and in Mismeasure

i think mismeasure is a piece of trash to be frank. gould was being a total propagandist.

There is just no value in these studies, and the results historically are always biased

there's value. psychometric tests correlate with abilities on other tasks. you might quibble with the causality, but there's an instrumental value. i.e., see jencks's 'black-white test gap,' the SAT actually predicts black college performance better than white college performance.

It strikes me as hubris to assume we are past these biases, and that somehow the research won't be used again to put down one race or another.

we aren't past biases. these sciences don't have the precision of physics. inferences are provisional and highly contingent.

It keeps on being stated as fact that cultural factors have somehow been eliminated,

where did i say this? i noted in an earlier comment that some cultures pretty obviously don't value repetition of tasks which might allow one to do well on tests which measure rational-abstraction capability. in other words, yes, i do think there are cultures which make people dumber or more intelligent than they might be.

Cultures in which there's never much need for explicit, high-concept cognition generally don't produce people who are practiced at that kind of thinking.

That doesn't mean that, when we find different performance levels between different cultures, we should be quick to attribute it to culture, no more than different levels between genetically distinguishable groups means we should attributed to genes.

When people from such a culture are putt in another, better-performing one, how well do their children do at abstract thought? Their grandchildren? etc. etc. That's one way we can begin to distinguish cultural and biological contributions to performance.