More Race and PC problems at Harvard

[Happy Ent]

Tangentially related blog post at gnxp/discovermagazine from today about African American genetic variation with really cool charts. (Also, the blog author, Razib is a big ASoIaF fan.)

[TerraPrime]

Maybe you relish viewing people that way, but I don't.

As a matter of fact, I do relish seeing people's ethnic and cultural heritage when I see them. I think it's rather high-handed of the people in the dominant culture to claim that one's status of skin color should not matter, when it does, every day. It must be nice to be able to afford to not see people's race, and "just" individuals, as if our skin color and how the rest of society react to that skin color had nothing to do with our individuality.

I shouldn't be surprised that you again misquote and misrepresent a comment despite having that inaccuracy pointed out to you before. That's your style.

I admit that you have made that argument before. I remain unconvinced. Others are free to come up with their own interpretation.

Re: Happy Ent

Just in case somebody actually still takes your claims at face value, here’s the closest thing to what you claim the paper says:

On the contrary:

On the basis of the present evidence, perhaps the genetic component must be

given greater weight and the environmental component correspondingly reduced.

In fact, Jensen’s (1998b, p. 443) latest statement of the hereditarian model,

termed the default hypothesis, is that genetic and cultural factors carry the exact

same weight in causing the mean Black–White difference in IQ as they do in

causing individual differences in IQ, about 80% genetic–20% environmental by

adulthood.

(p 279).

[Emo Targaryen]

And the question "Is there a genetic factor to black people's inferior intelligence" does not contain the statement "black people are less intelligent"? I am not sure what your point here is.

Did you follow the earlier pages of this discussion? Some of us are not convinced that this is a fact, based on different objections ranging from the biological concept of race as a useful framework to use in this context to historical biases in these types of measurements.

Well, as far as I am informed, black people (no matter how you define them) do indeed score lower on standard IQ tests, high school and college exams and the like. After all, if they didn't, there would be no need for any accusations of "institutional racism" and policies designed to fight it.

Did you read the Rushton-Jensen paper that Happy Ent linked? Those authors quite happily claim that genetics is the cause for a separation of 15 points in IQ score between black people and white people. That seems like a pretty exact and pin-point type explanation to me.

First, I understand they claim a significant influence of genetics, not a more-or-less single cause and their work is seen rather controversial. Second, if they really were able to undisputably pin it on genetics - well, then this debate is over anyway and we can conclude that the Harvard student was ostracized for telling the scientifically proven truth.

But Affirmative Action does not speak to IQ score or intelligence. It was not set up to address the issue of alleged disparate intelligence that divides along racial lines.

[...]

Is that how you think Affirmative Action works?

Maybe I'm wrong, but I understand AA to prescribe a certain minimum quota for each race to colleges and the like, so if their admission exams turn out an insufficient racial composition, they have to admit additional students solely by merit of their race. From the need for such policies I infer that black students would score lower on average than white ones in "colorblind" exams and that this phenomenon is seen as a product of institutional racism, from where I further infer that they would also score lower on IQ tests (as intelligence and academic success are somewhat correlated) and that this too is seen as a product of institutional racism.

[Former Lord of Winterfell]

Maybe I'm wrong, but I understand AA to prescribe a certain minimum quota for each race to colleges and the like, so if their admission exams turn out an insufficient racial composition, they have to admit additional students solely by merit of their race.

There are different types of "affirmative action". One type simply involves focusing on recruiting people as job candidates who fall into certain protected classes, but not considering their race, gender, etc., when actually making hiring decisions. That form of "affirmative action" is generally legal in all situations because the ultimate decision does not involve race.

At the other end of the spectrum is the hard quota type of "affirmative action" where you are required to place a certain % of people in a protected class in a given job. I suppose you could include in there systems that give a specific point "bonus" to candidates who fall into certain clases. That is more often illegal than not, although there are situations where it may be lawful.

ETA:

I don't want to suggest that these are fully settled issues, because they're not. There was a comparatively recent pair of Supreme Court cases involving the University of Michigan that addressed affirmative action. In one case, they found that awarding a specific number of points for being a minority was wrong. In the other, they said giving some sort of unquantified extra "consideration" to being a minority was, in fact, lawful. Obviously, that goes beyond mere extra efforts in recruiting candidates. I doubt its the last time the Supreme Court will take up the issue.

[TerraPrime]

Well, as far as I am informed, black people (no matter how you define them) do indeed score lower on standard IQ tests, high school and college exams and the like. After all, if they didn't, there would be no need for any accusations of "institutional racism" and policies designed to fight it.

There's a difference in standardized IQ tests, and standardized tests like SAT, MCAT, GRE, etc. The effect of discrimination can be seen in many areas, without having to rely on IQ test results.

First, I understand they claim a significant influence of genetics, not a more-or-less single cause and their work is seen rather controversial.

Their official stance is that it's 50%-50%, but if you read closely what they're saying, they often deviate from their own view (see the quoted paragraph I provided in response to Happy Ent).

Maybe I'm wrong, but I understand AA to prescribe a certain minimum quota for each race to colleges and the like, so if their admission exams turn out an insufficient racial composition, they have to admit additional students solely by merit of their race. From the need for such policies I infer that black students would score lower on average than white ones in "colorblind" exams and that this phenomenon is seen as a product of institutional racism, from where I further infer that they would also score lower on IQ tests (as intelligence and academic success are somewhat correlated) and that this too is seen as a product of institutional racism.

Your understanding of the Affirmative Action Plans is wrong.

The US Department of Labor, which oversees and regulates this, states:

For federal contractors and subcontractors, affirmative action must be taken by covered employers to recruit and advance qualified minorities, women, persons with disabilities, and covered veterans. Affirmative actions include training programs, outreach efforts, and other positive steps. These procedures should be incorporated into the company’s written personnel policies. Employers with written affirmative action programs must implement them, keep them on file and update them annually.

Notice that it does not say "employers must hire a fixed percentage of minorities." In fact, a quota system is, as far as I can tell, illegal.

As an example, my department right now does not have an African American faculty. We're not forced to hire one, under Affirmative Action. However, when we do look for a candidate to fill a position, we have to demonstrate to our Civil Rights Office on campus (which oversees the implementation of AAP) that we have taken positive steps (ergo, affirmative) to recruit from a diverse field of potential candidates. That's all we're required to do by AAP (well, our school needs a formal AAP that is approved, too, but to us, in implementation, that's all it's required).

[Happy Ent]

Their official stance is that it's 50%-50%, but if you read closely what they're saying, they often deviate from their own view (see the quoted paragraph I provided in response to Happy Ent).

Hm… you don't have to read anything "closely" — they aren't beating around the bush are are quite open and explicit.

Jensen has formulated the 80-20 hypothesis in a 1998 paper, but I'm not sure how strong he thinks the evidence for that is. What the Rushton–Jensen paper does is to compare the 50–50 hypothesis with the 0–100 hypothesis, and establishes that the former is a lot stronger in many dimensions. You, TP, routinely mischaracterise their stance as 100–0.

Jensen's 80–20 hypothesis makes some sense (because that's the within-group variance), so it's a good, bold target to put out. If it's extremely wrong, it should be easy to shoot down. That's how these things work. It's good science, no reason to call it to "deviate from their own view."

[Aemon Stark]

A few short points:

The continued insistence at dividing factors influencing a feature like intelligence into only two groups - "genetic" and "cultural" - is ludicrously outdated and resembles 1970s understandings of genetics. Genetic diversity exists within environmental variations that both affect and are affected by gene expression. When we get to the point of discussing "culture" or sociology, we are becoming very far removed from the genome itself. Failure to account for the considerable two-way interactions between genes and their environment is a fundamental flaw of the kind of research discussed here; it's all well and good to say that differences can be attributed 50:50 to "genetics" and "culture", but since these are not independent effects, a potentially very large source of variation will have been missed.

Regarding the Flynn effect, if IQ has not been stable in time (or, in fact, ever increasing) within the same populations, then it stands to reason that it is a poor measure for intelligence in any objective sense. Humans are not smarter than we were 50 or even 100 years ago... but that doesn't mean we're not better at IQ tests, assuming their results are as accurate as some think they are (Flynn discusses evidence why they're not, such as the fact that they are constantly re-normalized, but comparisons are and have been made between test outcomes with different standardizations).

Finally, assuming similar variation in the African American and White populations in the US, a difference of mean IQ of 15 points would actually imply a considerably greater proportion of mentally retarded and "borderline" individuals in the former population. I think this is implausible at best and probably not supported by any collateral data.

[Happy Ent]

Finally, assuming similar variation in the African American and White populations in the US, a difference of mean IQ of 15 points would actually imply a considerably greater proportion of mentally retarded and "borderline" individuals in the former population.

The actual numbers are indeed only twice to thrice as big. IQ 75 blacks function a lot better than IQ 75 whites. I don’t know much about this. (Nor do I care, “retardation”, like “obesity”, are semantic, discrete categories that I find toxic to quantitative reasoning. Also, I fail to see the usefulness of the single-variate IQ outside 1 or 2 standard deviations, it seems to me that predictive value should decrease when you move that far from the mean.) Maybe read more here: mental retardaion @ griffe du Lion . As far as I understand, the IQ 70-75 is a good indicator for retardation in Whites. I’m happy to learn more.

[Emo Targaryen]

As an example, my department right now does not have an African American faculty. We're not forced to hire one, under Affirmative Action. However, when we do look for a candidate to fill a position, we have to demonstrate to our Civil Rights Office on campus (which oversees the implementation of AAP) that we have taken positive steps (ergo, affirmative) to recruit from a diverse field of potential candidates. That's all we're required to do by AAP (well, our school needs a formal AAP that is approved, too, but to us, in implementation, that's all it's required).

Positive steps? So it means if no black candidates apply by themselves, you have to make some effort actively searching some, but you don't take their race into consideration when actually hiring them? No objections to this, but I was under the impression that some universities do indeed take the race of students into consideration when it comes to deciding admissions, be it by hard quota, bonus points or more fuzzy measures. I imagine this to be for the reason that students from ethnical minorities would underachieve if admission was solely based on SAT or similar tests.

[Former Lord of Winterfell]

Positive steps? So it means if no black candidates apply by themselves, you have to make some effort actively searching some, but you don't take their race into consideration when actually hiring them? No objections to this,

That is your standard federal contractor AAP.

but I was under the impression that some universities do indeed take the race of students into consideration when it comes to deciding admissions, be it by hard quota, bonus points or more fuzzy measures.

That happens as well, but those are usually voluntary programs, completely separate from the above-described mandatory AAP's required of all federal contractors.

[mcbigski]

Shit, I'm away from the board for a few days and miss this. And last night I got an invite to go to the Sox game with a buddy of mine who is graduating from Harvard Law in the next couple of weeks. Could have gotten the inside scoop.