I agree with Val that it's a good article. Let me comment on this part:

"In a pioneering study, the Florida State University psychologist K. Anders Ericsson and his colleagues asked violin students at a music academy to estimate the amount of time they had devoted to practice since they started playing. By age 20, the students whom the faculty nominated as the �best� players had accumulated an average of over 10,000 hours, compared with just under 8,000 hours for the �good� players and not even 5,000 hours for the least skilled."

The implication is that practicing the most made the best players the best. I think that is partially true, but the reverse causation also exists. Talented violinists will progress faster than less-talented ones given the same amount of practice, so the talented violinists find practice more rewarding and do it more. In school, high-IQ students get more
out of academic work than low-IQ students do, so they may spend more time on it. Studying calculus or closely reading a Jane Austen novel does not reward even average-IQ people, so they rationally avoid doing so.

The article discusses the predictive ability of the SAT and correctly states that it is a quasi-IQ test. Because of large group differences found on the SAT (also found on the ACT, on Advanced Placment tests, on NCLB-mandated achievement tests, and on the NAEP) , most writers on education don't want to think about the g-loading of standardized tests.

"To see what is in front of one's nose needs a constant struggle." - George Orwell