<p>How do you determine “a sample of kids who are equally proficient in basic math skills”? You certainly can’t use school grades… first of all, some schools are a lot harder than others. An A in this school may be a C in another. </p>
<p>Even in the same school, you can’t really tell the difference between one who gets 98% and the other who gets 99%. Maybe the person who got 99% had his parents help on those really hard homework questions while the other didn’t. Also, maybe the kid who scored 100% on every test but didn’t do any homwork got a significantly lower mark. </p>
<p>It MAY be more accurate if you take a sample of students who gets fairly low grades in the SAME SCHOOL. In that case, I do believe that the scores will be quite simliar. </p>
<p>There may be different factors, like what type of questions were asked. For example, I scored really low on the AMC 12 this year because most of the quetsions were on geometry. But that’s because I didn’t learn that much about geometry, and if I did, I would’ve probably done really well. </p>
<p>In conclusion, I think SAT is purely a measurement of knowledge. However, how easily you obtain that knowledge depends on these so called “other capacity”.</p>