|
Main
Menu |
« Previous Entry (older): Taking Criticism . . . and improving the profession.
» Next Entry (newer): Fisking Mitch Coleman
1692
Bias and the SATFebruary 11, 2004One little paragraph in a Cornell Daily Sun article on the SAT quoted in a post by Kimberly Swygert really got us thinking. A common complaint about the SAT is that it is biased against minorities and the poor.This might be, oh, because minorities and the poor do worse, on average, than other test-takers. An open-and-shut case for bias, no? Remember, correlation does not prove causation. The SAT could really be biased against minorities and the poor, sure. Or, it is simply accurately measuring skills in which, on average, students in low-income or minority areas (such as our inner cities) tend to be less well versed. Since it is uncomfortable to think that schools are underserving many students, especially students that need our help the most, then many folks chalk it up to that "mean ole biased SAT." This condemnation is often heard on college campuses, where ironically, research has consistently shown that the test accurately predicts college students' grades.Now here's a correlation we can stand behind: Do well on the SAT, and also get good grades in college. Do poorly on the SAT, and do poorly in college. Again, there could be a conspiracy afoot, such as when a professor is assigning grades. He could look up the students' SAT scores and assign grades that way, or he could grade students on how well they've done in class. So just maybe the SAT is accurately measuring academic skills. In the case of minorities, the SAT actually overpredicts college grades slightly, on average. Were it biased against them, their SAT scores would be significantly lower than actual school performance.We have one possible explanation: Affirmative Action. The article indicates that the SAT is an accurate predictor of college grades, then says that it overpredicts minorities' grades. One possible explanation is rampant racism all across the country, where minority students doing the same level of scholarship get graded down because of their race. A far more likely explanation is that many minority students have been accepted into colleges and universities where they aren't academically well matched. This is no ding on minority students, believe us. We can say, without pride, that we were not cut out for Harvard or MIT (although we were happy when a classmate was accepted into the engineering program at MIT). We ended up going to a state school. Under Affirmative Action, we are very confident that both of those elite schools would have accepted us, based upon GPA and SAT scores. Where we would have flunked out, in all likelihood. We think very highly of the SAT (except for the fact that they permit calculators even though the computations are not difficult). The fact that it overpredicts the college grades of minorities is one more reason for colleges to use race-neutral admissions policies. For more qualified minority applicants, let's fix our K-12 schools. Posted by ceb into Higher Education
, Testing & Grading
| ↑ top ↑ | « previous entry | next entry » | ReformK12 home Comments
Irony of ironies, the data seem to suggest that the SAT overpredicts first year college success for minorities. Yes, disadvantage minorities score lower than others on the test, but the test actually predicts better performance than is seen. By these data, the SAT could be eliminated or devalued for this group of students because it is biased toward higher readiness. Mike McKeown February 15, 2004 07:40 PMMike, ". . . biased toward higher readiness." Hmmmm. So I guess that settles it. The SAT is biased, after all. I shall inform the College Board promptly. ;-) chett February 15, 2004 08:57 PM |