Russell sees an apparent contradiction:
I was wondering if Peter Charnley could clear something up for me? He states that a new British study shows that men outnumber women at higher I.Q. levels. How does this conclusion mesh with the results of widely tested entrance exams like the SAT which show some difference between men and women on the math portion of the test, but nothing nearly as dramatic as suggested by the recent British I.Q. test. For instance, boys out number women by 2:1 in the higher reaches of the SAT math section (scores over 700). If this new study were true I would expect a much greater difference in male/female numbers than 2:1.
Renee gave it a good effort, but the answer is actually much simpler. The SAT is no longer an IQ test and it has not been for over a decade. In 1994, the tests were changed to focus on measuring achievement instead of raw cognitive ability, which is why Mensa will no longer accept SAT scores as evidence of the 132 IQ that is the required minimum on most of the IQ tests it accepts.
There is, of course, still some correlation between SAT success and IQ, it simply isn’t particularly strong.
“In the June 2004 issue of Psychological Science, Meredith Frey and Douglas Detterman of Case Western Reserve reported… that the mean correlation between SAT scores and IQ scores was .76.”