Cornell University psychologists Stephen Ceci and Wendy Williams defend the study of group difference in IQ against Steven Rose in today’s Nature. Ceci and Williams’ view is pretty much my own. They argue that, as often as not, claims about group differences in innate cognitive ability can be proven to be untrue. Actually debating the issue, and discussing the best new research, does more to defuse harmful misconceptions than a censorious egalitarian refusal to even publicly contemplate natural inequalities. Moreover, an ethos of punishing political correctness is inconsistent with the norms of open inquiry upon which communities of scientific discovery depend to produce new knowledge.
Of the James Watson imbroglio, Ceci and Williams write:
Watson’s first assertion could be read as scientifically supported: black Africans’ IQ scores are lower than those of white Europeans. But Watson’s use of ‘intelligence’ was interpreted as meaning ‘intrinsic cognitive ability’, ignoring how unfamiliarity with testing format, low quality of schooling, or poor health might depress IQ scores. There have been analyses showing average national IQs for sub-Saharan Africa to be approximately 30 points lower than average IQs for predominantly white European nations, and drawing a racial conclusion from those results1, 2. A refutation of these analyses would provide an opportunity to advance understanding. Sadly, although these analyses can be refuted, as we and others have done3, most of those who scorned Watson never knew they existed.
They go on to discuss interesting new work that is blowing away old stereotypes about the mathematical abilities of women:
Regarding gender, no one now claims women are unable to excel at complex maths: 48% of US mathematics majors are female, and women earn higher maths grades than men throughout schooling5. The maths gender gap among the top 0.01% of students, which 30 years ago favoured males 13-to-1, now favours males only 2.8-to-1 (ref. 5). Some nation’s women (including those in Singapore and Japan) outscore US males on maths tests by an amount far larger than the gender gap within the United States5.
So, vigorous debate has resulted in great progress in our understanding, and more breakthroughs will come — if we allow free speech in science.
The fact that that the gap between males and females at the top .01% has narrowed so dramatically so quickly is extremely strong evidence for the claim that prior gaps were a reflection of widespread sexist assumptions about female capacities. The clear implication of this kind of rapid convergence is that it is mistaken to take today’s snapshot of the distributions of various cognitive skills as indicative of innate or natural difference in capacities. Given a little more time, and little more erosion in sexist social expectations, we might find that females are as good or better than males at things like math — long thought to be a special male strengths. And the same argument applies just as well to preferences. One can’t take current patterns of occupational self-selection as dispositive evidence of “natural” differences in men’s and women’s preferences given the rapid past and current shifts in the gender composition of many occupations.
Here’s Cato Unbound‘s debate on the science and politics of IQ, featuring James Flynn, Stephen Ceci, Linda Gottfredson, and Eric Turkheimer.
It's a good piece, but the reference to Rousseau is incorrect. He certainly did not hold that view.
I certainly agree that this research should not be censored. But, I don't know many people who don't have a preference for what they'd like it to conclude.
I suspect that Will would prefer that it shows natural differences to be insignificant.
I would like to claim that I'm completely indifferent and just want to know the truth, but I think I'd kind of like to see the other result (clear differences, but small ones). Not because I'd like to support discriminatory policies, but because I've got a contrarian streak and would like to see PC zealots frustrated by evidence that nature doesn't conform to their ideology.
What's annoying here is the poorly hidden [if hidden at all] prior that a “good” result is that there is no difference at all in cognitive abilities between groups. It's quite similar to the bias in climate studies right now, where the one good and true outcome is pre-determined, and politicization creeps into almost every interpretation of research data. Such fuzzy-mindedness is puzzling from this blogger, actually. Eager to please a certain constituency, I would surmise.
we might find that females are as good or better than males at things like math
We might also find that women are better than men at math in general, but that men continue to be overrepresented at the top end due to slightly greater variation (and/or other traits like a greater tendency to monomania).
Other than that, I agree with Mesa. It's very nice to say *in general* that 'Error cannot be very dangerous so long as truth is free to combat it'; it's quite different to be cute and take this stance in a specific situation, where 'truth' and 'error' have already been assigned their sides.
I am not impressed with the math skills of those criticizing Wilkinson
Wilkinson: “They argue that, _as often as not_, claims about group differences in innate cognitive ability can be proven to be untrue.” [My emphasis]
How can this be read as being consistent with:
“…Will would prefer that it shows natural differences to be insignificant”
or
“What's annoying here is the poorly hidden [if hidden at all] prior that a “good” result is that there is no difference at all in cognitive abilities between groups”
?
I am not a card-carrying Bayesian, but I read “as often as not” as 50/50, a perfectly reasonable “prior” for such a broad statement. It would be peculiar if all early IQ research had to be thrown out as hopelessly biased, but it would also be peculiar if absolutely none of it was later found to be needing revision.
Will's disappointment with the stymieing of free speech on IQ research is rather humorous (as is the further excoriation by the Cornell profs) because that's exactly what's being done to IQ research everywhere. The respected professors see that the only use in doing IQ research is to refute these prejudiced claims in innate differences. The facts on the ground, however, are startlingly different. While it's true there are ample environmental causes of IQ depression amongst the poorest in the world, it is intellectually dishonest to attribute the entirety of the black-white IQ gap to said factors. When one controls for socioeconomic status, American Blacks still under perform American Whites by nearly a whole standard deviation, a result that remains true up and down the socioeconomic ladder (meaning that regardless of the socioeconomic decile you choose, whites will still outscore blacks by around 10 IQ points). One can assume that American Blacks in the top socioeconomic decile do not suffer from the pernicious environmental issues that our dear professors so readily mentioned (things like malnutrition or significantly poorer schooling).
And with regards to the claims about women and mathematics, all I have to say is that I generally buy the claims made by many of those in the environmental construction crowd, but with one central caveat. The smaller normal distribution curve for women with regards to IQ is undermined as a deterministic parameter by two key factors: women tend to over-perform on tests of conscientiousness, which is nearly as important as g in predicting success in any given field, and when one compares across races and ethnicities the differences become wholly insignificant. Amongst those with IQs in the top quartile, Asian women consistently outscore White Men on tests of mathematical reasoning, but at the same time, they consistently score lower than them on tests of verbal reasoning. But the reverse is true when compared to Asian Men. Asian Women score higher then them in tests of verbal reasoning, but worse on tests of mathematical reasoning. This would suggest that, although there exists some sexual IQ differences, they pale in comparison to other, more significant, factors.
The truths to which these fine professors attest to may not exist. Then what of their defense of the open exchange of ideas? They only talk of the discussion as if, prima facie, the more politically correct response is also the scientifically verifiable one. What if it isn’t? Would they (could they) support the discussion now? Based on my interaction with those who partake in the academic lifestyle, I’d like to think so, but I fear they would not.
And Will, I’m not going to let you get away with your statement on modern intellectual convergence and imply that things will continue to trend onward as it has. While the Flynn affect has shrunken the gaps somewhat in many countries, it has begun to reverse itself in two nations, England and Japan. The suggestion some have here is that the Flynn affect is more a reflection of opportunity gaps between the have’s and the have-not’s (whether that is men and women, rich and poor, or white and black), and as we have become wealthier and educational opportunities have become open to all, the effect will slowly begin to recede. There is some suggestion that the most important environmental stimulus needed for the increase in IQ is at least a decade of structured education, and now that that is more universal then ever before, the gaps in IQ will become more and more stable. It has already happened in some areas of the country and may spread further.
Other than that, congratulations on a fantastic blog. I love reading (and commenting to I guess) on such intellectually stimulating blogs, and I hope this is first of many I shall do.
The “…” was “I suspect that.” It was just a guess based on the tone and content of lots of posts; not just this one. It was certainly not an assertion or even inference from the comments in this post. It could easily be wrong.
My sense that people have preferences about what results they'd like to come out of this research comes almost completely from discussions outside of this post and blog.
May want to look at some contrary views on the math gap.
http://www.lagriffedulion.f2s.com/math2.htm
<I”When one controls for socioeconomic status, American Blacks still under perform American Whites by nearly a whole standard deviation”
Interesting. So, relative to whites, blacks are extremely high achievers, doing much more with what they've got than white people. Black persons with IQ's of 84 earn what white's of average IQ earn, while retarded blacks earn what whites in the low average range earn. And blacks with merely average IQ's earn what white's with IQ's of 116 earn.
How do they do it? Maybe IQ isn't as robust predictor of success as it's made out to be by bloggerss.
all this is interesting, but old news of course… “it's” all culture — we already knew that. no two groups of people who have distinct cultures will choose the same professions or excel in the same subjects at school or do the same on an exam… this finding is old hat. anthropologists have known this for half a century…
http://firelarrysummersnow.blogspot.com/
It's very nice to say *in general* that 'Error cannot be very dangerous so long as truth is free to combat it'; it's quite different to be cute and take this stance in a specific situation, where 'truth' and 'error' have already been assigned their sides.
Especially when the most comprehensive and exhaustive research on the matter thus far strongly indicates this is not an argument “the side of the angels” is going to be winning. In fact, if this was a less politically sensitive subject, it's a safe bet that this would largely be considered a settled matter based on the evidence as it stands. But, considering the social and political implications, one suspects no evidence, no matter how conclusive, will ever be considered adequate in some quarters.
Why are you linking to a ten year-old book? Read Ceci. Read Turkheimer.
Self-published?
““more erosion in sexist social expectations“
Not just expectations, actual barriers – actual barriers, says the woman who was told by her high school “guidance counselor” that she couldn't take Algebra II because 1 – “the class was full” and 2 – “she wouldn't want to be the only girl there.” This was in 1978, people! I went on to study conic sections from Apollonius (ancient Greek!).
Mwah-hah-hah. You won't be able to use entrenched social power to protect you from intellectual competition much longer, you heterogametes you.
But seriously, the power of testosterone in the womb to shape brains is amazing. Due to these effects, I will never be surprised if men on the far extreme of math ability continue to outnumber women at that extreme, just because of this factor.
Men and women are different in some ways. Vive la difference! That doesn't prevent us from establishing a just society, however.
Back to this dorky g factor. I just don't believe in it. No one can agree on how to describe this puppy, no one can agree how to measure it, no one can predict with it — it's just a unicorn, folks, give it up. What the hell is scientific about it?
Those stupid tests are completely biased in numerous ways, they just are. African-Americans are disadvantaged by them, while people of my ethnicity are favored by them. They should all be scrapped. If you believe otherwise, then you're stuck believing I'm a near-genius and the people of my ethnicity are on the whole the smartest people in the world. Which just ain't so. Prima facie, Nobel prizes or not.
Why are so many (guys) attached to this idea so strongly? Isn't that suspicious?
jlerner: “When one controls for socioeconomic status, American Blacks still under perform American Whites by nearly a whole standard deviation”
Ben Wechsler IV: “Interesting. So, relative to whites, blacks are extremely high achievers, doing much more with what they've got than white people.”
No, because the first statement is in reference to background factors, not earned income. Controlling for IQ closes most of the B-W wage gap. So blacks and whites matched for IQ have similar occupational returns.
On the other hand, we see very different IQ outcomes for the children of blacks and whites of matched economic backgrounds. Black children from homes in the highest socioeconomic bracket have lower IQ scores than white children from homes in the lowest socioeconomic bracket. And their adult income reflects their IQ score virtually regardless of race or class background.
So this could be interpreted either as blacks underperfoming given initial socioeconomic advantages, or neither blacks or whites performing any differently than you would expect give initial IQ scores.
“Self-published?”
Figure 8 is the best piece of evidence offered on this question so far.
La Griffe is no piker, and the PISA numbers are all available to check online. Williams study isn't published until next month, so we can't evaluate it's methodology or know what data set was used. But it's hard to compete with the PISA data if we are asking how much cultural differences matter.
This Turkheimer?
You are joking, aren't you?
Hello,my name is robyn King I am a 24-year-old female living in Los Angeles,California The United States Of America and yes I think and would say that all 100 percent both males and females japanese people all over the world have the most highest I.Q. ever recorded and that all of them score 228 I.Q.'s all over the world.The reason why I say this is because 100 percent of them get and are straight A+ students and straight A students all through school and college all through their lives from generations after generations and they are always really very extremely smart in school and education.They have an extremely very high excellent education rate and are the most smartest race of people in the world and have the very top rate of their education.
Hello,my name is Robyn King I am a 24-year-old female living in Los Angeles,California The United States Of America I think and know that all japanese people and all of the oriental asian race of people are 99.9 percent near perfect both males and females and for having extremely very high I.Q.s like that and I wish that they were perfect!
But in 1978 he changed from commercial illustration to historical art. In 1997, Dietz won the american flag wallpaper Stanley Wanlass Award for 'excellence in strength of design and composition echoing the spirit of the automobile' at the Meadow Brook Concours d'Elegance.
But in 1978 he changed from commercial illustration to historical art. In 1997, Dietz won the american flag wallpaper Stanley Wanlass Award for 'excellence in strength of design and composition echoing the spirit of the automobile' at the Meadow Brook Concours d'Elegance.
Pingback: The American flag clip art, text has a light effect « American flag wallpaper