With all this talk about IQ, I was tempted to take one of those web IQ tests. And of course, what comes to the fore is the very, very obvious way in which racism could affect IQ: by affecting hard you try when you get to a difficult problem.
I hate those stupid "rotate the object" questions; I'm not good at them, and they're boring. So when confronted with one, I was tempted to give up. But I thought to myself: you know you can do this, since you've done it before, and besides, I want to see what kind of number it pops out.
Consider all of the environmentally imparted values in that thought:
1) IQ tests are important and worth trying on
2) You are a smart person who can solve problems
3) You have solved this kind of problem before, and therefore can do so again.
Then think of someone who has grown up in a home, a neighborhood, or a school district where these tests aren't important and worth trying to excel at, where you perhaps internalize a belief that members of your racial group aren't good at these tests--or even shouldn't be good at these tests--and where you may not have encountered these kinds of problems before, so you don't know that you can solve them.
It's not at all hard to imagine that a black person with exactly the same capability to solve the problem as the white person sitting next to them might nonetheless fail to solve the problem. And of course, played out on a sufficiently broad scale, this will look like a heritable group difference, since almost all black kids are raised by black families--and even where they aren't, are going to be treated as black by everyone they meet, and internalize whatever messages our culture sends about blackness. There's also some evidence that people perform worse on tests when they are told that their gender or racial group doesn't do well on the test--and are black kids ever told anything else?
So yes, I think that IQ tests could easily widely overstate the intractability of IQ, particularly for intergroup distributions. I'm not sure where that leaves us, since it's hard to alter society in the ways that this analysis suggests. But at least it's a little more hopeful than "They're just born stupid."
Update Edited to correct the weird wording
This is a testable hypothesis. Just give a test where you pay people 1 day's salary for every question that they get correct. Integrate the IQ questions with trivia and other questions so it doesn't seem like an IQ test.
I agree that IQ tests are not reliable indicators of intelligence. All kinds of things can affect one's performance on such tests--including practice at taking them. But who thinks that Africa's troubles have resulted from genetic inferiority? Is there any serious person who makes such a claim?
Many things have gone wrong in Africa. Geography and disease have certainly played their part. But the main reason things have gone so badly in the last century, aside from the many tribal conflicts, is that socialism and communism took hold there. While the US, Europe and East Asia embraced markets, Latin America adopted protectionism, central planning and resdistribution.
I've found your work awful and uninformed, but this is ridiculous. You write as though there's been a desperate search for an explanation for black Americans' performance in standardized tests other than "they're just born stupid" and you have, just now, provided the first viable alternative.
If this doesn't lead to a stern discussion with the Atlantic's editors they've got some 'splaining to do.
run this through the gender machine and you get a similar result... if someone tells girls that "Math is hard..." you think it is that way.
Getting back to the idea that various test for broad ranges of people are going to be biased based on the inherent interest a group has for testing.... That would be the reason a test writer or company would write a test with specificity of some sort towards that group. You know the catch as well as I. Who decides what bias and how does it all apply on the supposedly level playing field?
Odd as it sounds, it may well be easier and certainly better in general if there were a push to help individuals of a group, racial, gender or other to see the value of learning and testing. Rather than trying to tailor every test to every groups biases.
It's a simplistic approach, but I think we lost sight of it a long time ago.
Kudos to you Megan. You know you've made it when people who ostensibly hate your work are still reading your blog.
That should read "...Latin America and Africa adopted protectionism..."
A lot of people (e. g. Jeff Sachs, Angelina Jolie) want to make Africa's trouble look like nothing more than misfortune. But it was not just chance that brought economic disaster to Uganda under Idi Amin, nor to Ghana under Nkrumah. Nor is it mere chance that has brought reasonably good economic growth to South Africa, even while Zimbabwe suffers. Policy choices have made the difference.
Megan's hypothesis has been tested. Psychologists do research on "stereotype threats" and have found that simply reminding test takers, even very subtly, of stereotypes about their cognitive abilities can affect how they perform.
Megan,
These symptoms that you mention are culturally based, not race based. I know that you have been held up in Yankee land most of your life, but you need to come down and meet some strains of W.T. down here in the south. The inability to break out of your life script, due to environment, lack of help, or laziness does not care what color your skin is. I have seen (smart) members of my extended family willfully exclude themselves from good middle class lives due to poor choices, and even SUPERSTITION.
So, I guess what I am saying is that you need to rent Urban Cowboy (my family history), and expand your data set.
-Donut, Y'all
ps. After you see Urban Cowboy, make sure to remember that the smart, educated Texans can sometimes hide behind those personas, and you never know what happened until after you have already signed the contract.
Megan (and Sasha):
There's an interesting paper on ethnic stereotypes and academic performance here:
http://64.233.169.104/search?q=cache:2OvlHFbOmHkJ:www.sitemaker.umich.edu/culture.self/files/altschul__oyserman__bybee__2006_.pdf+%22+Racial-Ethnic+Identity+in+Mid-Adolescence:+Content+and+Change+as+Predictors+of+Academic+Achievement%22&hl=en&ct=clnk&cd=2&gl=us
The authors suggest that being reminded of group identity tends to reinforce behaviors consistent with stereotypes, i.e. Asian-American students in an environment that emphasizes their Asianness will tend to feel pressured to perform academically, but black students in an environment that emphasizes their racial identity over other factors will tend to conform to low-achieving stereotypes.
Mike, obviously I do not believe that I, Megan McArdle, have just discovered this astonishing new theory. I was using a personal example to illustrate the point to readers who may not have thought about it.
Who cares about IQ or other measures of potential which will always be controversial? And anyway, you don't have to be a genius to hold down a job and put a roof over your head, do you? So what do a few IQ points matter? How about more objective measures of actual life success - employment, home ownership, income, etc? Do those mean anything? OK they do. Now you have to explain differences there. How do you do that? Well I guess you look for root causes of those differences. If you don't like the idea that potential may be different then perhaps opportunities are different. Well let's look at economic circumstances, or school quality and control for that. Well, we still find differences, now what? Blaming cultural attitudes towards family formation and education isn't going to cut it in the academy. So we are left with... "racism". "Racism" is good, because it blames the successful for the failures of the unsuccessful. Also, there's nothing to be done about it except complaining and hand-wringing. We can all be "concerned" and "saddened" Maybe we need another 100 years for that crutch to disappear.
This is one of those dichotomies that gives you no choice but to hold an invidious racial view. "They're born stupid" or "they just, culturally, don't try very hard." I'm not sure I want to go either place.
Are the terms "stupid" or "apathetic" applicable to whole populations. I failed an organic chemistry class in college because I didn't try very hard, and that was stupid. But some number of students in any organic chemistry class are destined not to do well. And between, say, Harvard and Brown, the number of students who don't pass organic chemistry may vary reliably, and perhaps intractably. Does that mean the student body at Brown is stupider than the student body at Harvard? If not, is the necessary conclusion that the student body at Brown is more apathetic? I don't think you can draw either conclusion. All you can say is that more students at Harvard pass organic chemistry than at Brown. An accurate model of why that's true is bound to be more complex than just taking my pathetic organic chemistry story and multiplying it by the size of the population.
Re: "genetic inferiority?"
It is very unPC of me but I want to know if there is ANY biological basis to social inequality. The answer to that question will very much inform the measures I will support to maintain social cohesion.
There are many pathways for heritable information to be transmitted. DNA is only one, heritable proteins coding for birth weight in rodents allow for rapid (generation to generation) adaptation to environmental conditions for which the process of encoding in DNA via natural selection is too slow. Why not then cultural transmission of predispositions to behaviors?
Nor am I implying that behaviors which lead to lower IQ scores (for instance) are maladaptive. A cultural response to racism could lead to both high academic achievement in Jews and low academic achievement among the Roma (Gypsies).
The reluctance, expressed in this thread to even consider the possibility of biological/genetic/cultural mechanisms for suboptimal performance betrays a fear that at the base that these issues are intractable. If the basis of African America under performance on IQ exams is genetic, then the attendant social problems can not be addressed in less than the centuries that natural selection needs to work with. Social cohesion would then demand that we simply pretend the problem does not exist in much the same way we ignore other unpleasant social realities. On the other hand, if the problem turns out to be a matter of social adaptation, or simple gestational and early childhood nutrition then we would be well advised to KNOW that fact so that we can address the cause.
Stone Mao
The rotating shapes problem is an example of (in my opinion, not supported by links or any facts, just my anecdotal life experience and no one has to agree with it) "female" thinking. Males have a better understanding of three-dimensional abstract concepts than women. This makes men better (as a group) in science and engineering and has nothing to do with the psychobable excuses.
It is also my experience that the brains inside males of different races are wired slightly differently. IQ tests a cultural desire for what is considered intelligence. Blacks are wired to be a bit lower than whites who are wired to be a bit lower than asians. This indicates that the IQ test is not a perfect evaluation of true objective intelligence.
In addition, IQ is not always an indicator of success or common sense. While everyone sniggles about poor Africa, what about Asia? With a culture of superior IQ, why does most of Asia live in crowded, polluted, rat cages ruled by dictators. Doesn't sound like a race of super-brains to me.
Everyone and every group has some deviation from the norm that includes special talents and deficiencies. The distribution of wealth in the world is a largely a function of geography and the march of history.
Donut's PWT relatives are a good example of devolution caused by a successful society. Since all of the basics are squared away and taken care of by the society machine, individuals may not feel the heat to make it happen.
Slightly off topic question. Does anyone have any idea why all of the new and improved humanoids develop in Africa first? This is even after migration into Europe. If one were to select the one area of the world today and predict where the next better model beyond sapiens-sapiens will come from, it would be hard to select Africa. Is this issue related to differences in IQ scores? I don't know.
You say,
Then think of someone who has grown up in a home, a neighborhood, or a school district where these tests aren't important and worth trying to excel at, where you perhaps internalize a belief that members of your racial group aren't good at these tests
And later say,
black kids ... internalize whatever messages our culture sends about blackness.
By "our culture," I assume mean "our white-dominated culture." But then the two statements are very different.
The second is what white people say to black people. The first is what black people say to black people.
The second historically said, "You're black; you're inferior." Today the official line, just about unanimous in schools and the media, churches, etc., is "Of course, black people can do as much as anyone."
Today, it is the first that says, these tests aren't important, they aren't worth trying to excel at, you won't do well (after all, white people made them up).
Mike, Sasha,
I think this once again goes to Meg's general laziness at facts and desire to be a journalist without having to do the work of a journalist.
Um, yes, Justin indeed it has been tested; that was the "evidence" I was talking about. I'm not sure how this makes me lazy. The generalized group reinforcement research, on the other hand, is new to me.
We are all individuals. Here is a thought question for you.
"How well do I represent the human race? All members of the human race, good and bad, etc."
If your explanation were a strong factor in IQ test performance, we would expect to see a smaller gap in young children than in teenagers and adults, as people become aware of their stereotypes. It's hard to imagine cultural stereotypes having much power over a two or three year old child, however performance on tests of children at that age are predictive of performance on adult IQ tests.
What scares me is that the question - is there any genetic component to IQ differences across races - is ultimately a scientific one. It seems that many people are making the mistake of applying moral aphorisms to this question. We don't want to confuse what we think should be the case in a perfect world with what is the case in the world we inhabit. I'm not an expert, so I'll remain agnostic on the answer, but let's stop pretending that if we just wish really hard, we can change the result.
I watched Everybody Hates Chris last week too.
by affecting hard you try when you get to a difficult problem.
Perhaps you want a "how" before "hard"?
Meg, the research in the field goes WAY further than what you ACTUALLY stated:
"There's also some evidence that people perform worse on tests when they are told that their gender or racial group doesn't do well on the test--and are black kids ever told anything else?"
Almost as irritating as your desire to play journalist but not work as a journalist is your desire to defend your errors by claiming that you said something you didn't, or that you meant something against all plausible forms of interpretation. You are too good of a writer to try to claim now that this one sentence was meant to show that you explored the research in the field - even briefly - before posting. Far more likely is that you just remembered some survey one of your friends told you about in a coffee shop and added it for magic "research-like" effect.
Megan, please explain how American social culture depresses IQ scores of Africans in Africa.
I'm waiting...
Oh, well, thanks anyway.
In other news, please note that you have reinvented the IQ literature's "Factor X," which mysteriously affects between-group but not within-group IQ differences (e.g., affecting the IQ's of black but not deaf people).
The problem with Factor X-- Dickens and Flynn's admirable attempt to banish it not withstanding-- isn't just that it seems independent of any mode of action you can hypothesize for it, but that it also fails to explain trans-generational reversion to the mean. Dickens and Flynn fail to explain that too-- I linked their paper so you can see that even their theory doesn't account for the fact (easily explained by the hypothesis that there is a substantial genetic component to IQ) that black children from rich families average SAT scores (proxy for IQ) even lower than white children from poor families.
The biggest problem with talking about race and IQ is that the whole subject is so controversial that even stating what is known right now generates such a strong emotional response that discussion of what consequences or policy prescriptions follow from those facts is effectively impossible.
Look, for example, at The Bell Curve. A lot of that book was rehashing of the research that shows that yes, American blacks have significantly lower results, on average, than American whites, and that people with lower IQs tend to have less desirable life outcomes, on average, than those with higher IQs. But that's only half the book. The rest of the book asks "What should be done", and provides some answers.
Some of the answers attack the problem at the root - getting lead out of children's environments will raise children's IQs (or prevent them from being lowered). But that only helps future people. Murray and Herrenstein offer other proposals to help ameliorate the effects of the IQ differential on adults now living and working. Murray endorses some level of affirmative action, though thinks that the bonus should be well-defined and limited, as opposed to the open-ended quota-hunt which it has become.
I thought ALL people were "born stupid". Although some mothers seem convinced otherwise.
Why does Megan hate black people?
Megan, I think IQ research is a lot more scientific than you give it credit for (ie you might realize that online IQ tests are for entertainment purposes only). I also think your naive to think that there is some obvious non-genetic factor that you could simply think up that explains the black-white IQ gap. This gap has been thoroughly studied, by a wide variety of organizations (Military, SAT, etc...) and by a lot of people who wish that it would go away.
Also if your mechanism were true, it would be that blacks gave up on the marginal questions, not the hard questions** (which by definition few people would answer correctly). It takes more of a leap to say they gave up on question they could just barely answer than one that was "hard". Yes it could happen in individual instances, but enough so that in aggregate it shifts the average a full SD of the white distribution? That seems pretty amazing.
**If higher IQ people don't get a question correct more often lower IQ people, then the question isn't correlated to intelligence, and not proper for an IQ test. (If that definition seem circular, try writing a question that high IQ (or whatever proxy you want to use) people don't answer correctly more often than low IQ people. You will find that is either knowledge based, ridiculously hard or ridiculously easy.
Steve Sailer has already discussed the exact problem you mentioned. The Army gives out intelligence tests but was obtaining scores that seemed too low, with the reason being that people got discouraged. So they changed the tests to start giving easier questions when people do poorly.
The real issue with whether or not an IQ test is "good" seems to me to be its predictive power. They have quite a bit of it. If people get discouraged by hard problems perhaps that something we WANT the tests to reflect.
Has it occurred to you during your "research" for this post that these psychometricians who spend a lifetime thinking about these things may have come up with clever ways to correct for just this type of bias?
http://www.opinionjournal.com/extra/?id=110007391
Search "A concrete example..." for an example of the sorts of test I reference.
This is genuinely racist. Any regret I ever had for starting firemeganmcardle.blogspot.com has been permanently washed away by this post.
Blacks are not lazy.
Megan:
Millions of people have had their dreams of a military career shattered by being unable to score high enough on the military's AFQT IQ test, no matter how hard they tried. I know one young man who, after failing once, spent months at a special AFQT tutoring camp for avid would-be recruits run by the Army, He loved this taste of military life and was chosen Best Cadet by the sergeants. And then he still flunked the AFQT again. (By the way, he had a black father and a white mother, for whatever that's worth.)
During the easy recruiting years of 1992-2004, the military turned away virtually everybody who wanted to enlist who scored at the 30th percentile or below (a 92 IQ), and has only grudgingly let in a few more low-scorers since Iraq turned bloody.
The crucial points are that IQ tests have predictive validity, and they predict real world outcomes just as well for blacks as for whites.
This doesn't mean any IQ test can foretell an individual's life accurately, it just means that a reasonably short test can provide highly useful insights into how large groups of people divided up by IQ will perform on average.
That's why the U.S. military spends a fortune on IQ testing all applicants for enlistment. The military has done numerous studies showing that its IQ test (the AFQT) is a useful predictor of trainability and on-the-job performance across a wide range of military assignments. A large group of avionics technicians with a 110 average IQ will have fewer planes crash on them than a large group with a 95 average IQ.
To say that the AFQT is a racially unbiased predictor is to say that a black with an IQ test score of 95 will not perform better in the military on average than a white with a score of 95. It's all been researched endlessly by the Pentagon.
Now, it could be that the black recruit who scores a 95 really _could_ have done better on the test than the white recruit who scores 95, but that the black was held back by some mysterious X Factor. Likewise, maybe the black who scored 95 really could do a better job while on duty of keeping planes from crashing except for some Y Factor that's holding him back. But, from the military's point of view, the key is that he doesn't actually _do_ either. On average, the guy who scores a 95, whether he's black or white, keeps planes from crashing like a guy who scores 95, not like a guy, white or black, who scores 110.
As for whether the 0.97 standard deviation difference between whites and blacks when the AFQT was recalibrated on a nationally representative sample in 1998 is caused wholly by environment or by some mixture of environment and genetics, well, we'll know in a decade or two as the DNA research progresses. If you want to know sooner, you can advocate for more direct research funding of this topic.
In the meantime, the one thing we can know for sure at present is that a racial average IQ gap of some size will be around for some time, so we ought to get used to trying to think intelligently about the existence of this massive social fact.
In summary, may I suggest you put aside the whole topic of IQ until after you've done some more reading about the subject? You're getting publicly irritated that people who have worked harder than you on learning this complex topic aren't taking your ideas seriously. The problem is that they've heard them all before, they know all the responses, the counter-arguments, the counter-responses, and so on. At present, you don't. There are lots of people who could help you get started in reaching the next level of knowledge. All you have to do is ask.
Well and graciously said, Steve Sailer. Hopefully Megan will take your advice. She deserves credit for at least being willing to discuss the topic.
I've found your work awful and uninformed, but this is ridiculous. You write as though there's been a desperate search for an explanation for black Americans' performance in standardized tests other than "they're just born stupid" and you have, just now, provided the first viable alternative.
Woman-hater.
What scares me is that the question - is there any genetic component to IQ differences across races - is ultimately a scientific one. It seems that many people are making the mistake of applying moral aphorisms to this question. We don't want to confuse what we think should be the case in a perfect world with what is the case in the world we inhabit. I'm not an expert, so I'll remain agnostic on the answer, but let's stop pretending that if we just wish really hard, we can change the result.
If we want to determine whether there's a genetic component, than yes, we do have to worry about what would occur in a "perfect" world.
One can't simultaneously dismiss the attempts to control confounding factors as PC wishful thinking and still claim that resulting research would say anything about genetics.
This thread is an awesome illustration of "Unskilled and Unaware of It", and shows how dangerous people become if you give them one but only one statistics class.
Steve Sailor:
I don't read anything in her post that is inconsistent with anything in your post. Isn't the key point in both that IQ tests are measuring things other than "raw intellectual ability?" You
simply added "but that's still valuable," which is not something that I see Megan denying here.
Your basic argument has no racial dimension, for there are underclass/unencouraged individuals of all races/backgrounds.
Correcting for income and class doesn't eliminate the measured disparities, as far as I'm aware. That fact is profoundly disappointing to me.
The essential point is that, as I've said before:
- the current racial gaps in average IQ are massively important social fact swith ramifications for countless issues (e.g., immigration policy);
- and, whether or not they have a genetic component, they are not going away anytime soon. (It's been largely forgotten, but there was a concerted, explicit federal effort to narrow the IQ gap in the 1960s and 1970s, with few tangible results.)
Thus, the political class ought to take the racial gaps in average IQ into account when thinking about public policy, yet we absolutely do not, as, say, the ridiculous No Child Left Behind act shows. (Bob Herbert has a good column in the NYT today on how stupid NCLB is.)
Instead, those who actually have spent time studying the facts are demonized as "racists" and their expertise is ignored.
The pernicious side effect of bringing up tired, discredited old ideas like Megan is doing here is that it feeds into a self-reinforcing anti-reality cult.
The demonizers cling to the assumption that the thinnest thread of hope that the causes aren't genetic automatically means that the whole topic can be safely ignored because it will just go away Real Soon Now. If they get asked what ought to be done to close this 100% environmental gap, wel hear that "all" we have to is whatever pet cause they already promotes for other reasons, such as vouchers or higher pay for union teachers or get youth to listen to Mozart instead of rap or abolishing the Welfare State or whatever, with zero awareness that most of this stuff was studied in the past, and the reason you don't hear about it anymore is that most researchers gave up in despair.
Thus, the quality of elite discourse in the United States is completely hamstrung by the many of the most crucial facts being off limits for discussion. Perhaps that's better for all concern, but I happen to believe that the truth is better for humanity than lies, ignorance, and wishful thinking.
Megan, so after the test, what number did pop up?