<p>FastMed, there are two confounding factors that I have seen. The first is that the test can be trained for. This is consistent with what you state. The content of the test is from 8th, 9th and 10th grade, so it becomes more a matter of mistake avoidance than knowledge. My D took the SAT while taking precalc and was therefore at least two years from the material on the test. She raised her score on the retake 70 points while taking calculus by taking sample tests (2) out of a book and looking at where she was prone to make errors. The other tests can be taught as well, so they are not a pure test of ability. The other confounder is the tendency of some takers to choke. Most males release epinephrine when stressed (“This is a timed test, and the rest of your future depends on the result”) and therefore perform at a higher level on the SAT than they have in school. Many females release acetylcholine when stressed and perform at a lower level than they do in school.
I agree that the test is not totally useless. The colleges can use it to compare schools and as a very rough guide to ability, but is like cutting with a chain saw compared with cutting with a microtome. My D did not improve a standard deviation in 4 months in her ability to do math. She was a good student before, had the math teacher both years and idolized him.</p>