Yeah, I have trouble believing that there is good research supporting this assertion. I think that the author meant to distinguish between someone who is 1 in

Yeah, I have trouble believing that there is good research supporting this assertion. I think that the author meant to distinguish between someone who is 1 in 100 and someone who is 1 in 10,000. My problem here is that I suspect that it's stupid to use 7th grade SAT math scores to make this distinction. In fact, I doubt that any standardized test can make that distinction - I suspect that they top out around the 99%ile (actually, an 800/800 on the GRE math was only about 96%ile when I took it). When a test is tedious, irritating, and consists primarily of solving simple geometry and algebra problems and choosing answers from a multiple choice menu, and when the difference between 99.9%ile and 98.7%ile is a couple of missed answers, I suspect you're measuring noise.n I'm not saying the tests are useless, they're probably pretty good. But to identify why the few great mathematical minds come from one group or another? Seriously? The only real way to measure this kind of talent is to provide opportunities to learn, step back, wait 20 years, and see what they've done. As for the gender question, well I guess I'm just begging the question with my "opportunity to learn" statement, because the whole question is whether this imbalance is in itself strong evidence of differing opportunities for men and women. How do you measure something that is obscure and present in only a miniscule fraction of the population? Our indicators can tell us who is 1/100, maybe kinda. This whole approach is fubar.