The Obama effect: Have blacks in general suddenly become smarter?
The NYT said so last January and the academic study upon which that claim is based has recently become available online. So I suppose I should say a few words about the absurdity. The first thing to notice is that everybody else seems to think it is an absurdity too. They don't put it as bluntly as I do, though. What they say is "This finding will have to be repeated by others before we take it seriously" -- or words to that effect. And the reason why they say that is that Left-leaning social scientists have been labouring mightily for many decades at the task of getting black intellectual achievement up to white levels. And nothing that they try works. So to say that the election of Obama has suddenly closed that pesky gap is improbable to say the least.
I don't have access to the full academic article but what I see in the abstract immediately reminds me of "the dog that didn't bark" in the delightful Sherlock Holmes story The Silver Blaze. The research involved giving the same group of people the same test four times. Now that immediately puts into the mind of any psychometrician "The practice effect" and so one would expect some mention of how that problem was dealt with. But there is no such mention. When you give the same test to the same people on two different occasions, you find, for various reasons, that they get higher scores the second time around. That is the practice effect. And to give the same test to the same people not twice but four times sets all the alarm bells about the practice effect ringing.
So there are two ways in which the final (post-Obama-election) results reported could simply be an artifact (product) of the practice effect: 1) Everybody had got so good at the test by then that hardly anybody got anything wrong -- thus equalizing the scores for blacks and whites; 2). Maybe blacks worked harder than whites at figuring out where they went wrong on the first couple of occasions and for that reason alone got their scores up to white levels eventually.
The only way those two possibilities could be precluded would have been for the authors to use not the same test four times but four parallel forms of the same test, and there is no mention of that. Parallel forms have to be very carefully constructed to ensure that they DO give the same scores for the same people and that is so onerous that I have never seen more than two forms of any test made available.
So the entire study would seem to be methodologically naive and incapable of supporting its conclusions.
I might mention that the entire study is the latest variation in the absurd "stereotype threat" literature. The stereotype threat theory says that blacks do badly at tests because they think blacks do badly at tests. The initial "proof" of the theory arose from a study wherein psychologists made some black test-takers especially aware of their blackness while others did not have their blackness mentioned. The more aware blacks got worse results. But the unaware blacks still scored the usual amount below whites. So it showed, rather clearly, that awareness of blackness was NOT the cause of the black/white gap -- as the unaware blacks still did badly. Awareness of blackness can WORSEN black performance but unawareness cannot IMPROVE black performance. But to this day the theory is believed by most academics who refer to it. They still assert that awareness of being black is why blacks do badly. There are many other absurdities in the theory, one of which is that it seems to apply only to blacks and not to all minorities, but anybody who wants to look at the matter in detail should read here, here, here, here, here and here.
REFERENCE: David M. Marx, Sei Jin Kob and Ray A. Friedman (2009) "The “Obama Effect”: How a salient role model reduces race-based performance differences". Journal of Experimental Social Psychology, Article in Press.