A New Book Argues Against the SAT (bias against minorities, women asserted)

New York Times ^ | November 9, 2011 | REBECCA R. RUIZ

[Mr. Bird]

He characterizes it as a test that tends to favor white, male, upper income students with the means to prepare for it. Chang Young Chung, a statistical programmer, and Thomas J. Espenshade, a sociology professor, both at Princeton University

Irony Alert.

[Reeses]

It’s now Wake Forrest Gump University.

[Spok]

The truth is the SAT accurately predicts success in college. The correlation is indisputable.

[Hop A Long Cassidy]

I agree.

The guy’s book is on Amazon and you can read a chapter written by Charles Murray who makes interesting points that the author selects from. Apparently Murray says that the College Board’s SAT test doesn’t offer any more information than College Board’s achievement tests (math, English, etc.) which is not really a condemnation on the test itself. Do all colleges require the achievement tests?

Because of rampant cheating in schools (think Atlanta, GA), some forms of achievement tests are needed to evaluate candidates, and I think SAT scores work well. And, if they are not harmful, why is this guy so adamant against them.

If IQ’s were used by high schools, it would be easier to judge the success or failure of high schools. For example, it’s unfair to condemn a school along the Mexican border or in the ghetto because there are limits to what the students can do. Oh, that’s right, talking about race and IQ is a no-no.

[morphing libertarian]

Thanx. Posted on the Christmas Thread. Hoped it was satire.

[Boogieman]

I’d wager there is a cultural element to the differences as well, insomuch as parents usually need to push their kids to study and take practice tests, then retake the tests to get a better score. If the parents aren’t involved, then the kids will not do these things, and their scores will suffer.

[Atlas Sneezed]

Plenty of good points on this thread, but keep in mind this big one.

Standardized tests exist to PREDICT future performance (like success in college). Bias can easily be discovered if it turns out that one group is OVER-performing compared to what their test scores would have predicted. (This requires an honest measure of the later performance being predicted).

If blacks were kicking academic butt in college beyond what their scores would suggest, then the tests are probably biased.

I’ve seen no evidence of this.