Girl’s SAT scores are lower than boys because of bias in the questions, charges a Center for Women Policy Studies report. Nationally, boys score higher on 4 of the verbal questions and 17 of the math, and the fact that they do better is alone prima facie evidence, according to Phyllis Rosser, that the Educational Testing Service’s (ETS) SAT test is weighted in favor of the male sex.
What we should be asking, of course, is not just on what questions do boys do better, but why? In Rosser’s studied sample of about a thousand upper-middle-class New York area students, she found the boys did substantially better on 10 out of 60 math questions. Of those 10 questions, 3 were specifically about “boys'” enterprises, suggesting to her that the context of the questions, rather than the ability being tested, was adversely affecting girls’ scores. For those questions that seem to favor boys (or girls) simply because of the context in which they are put—figuring out a basketball score, for example—a valid argument can be made that those questions should be replaced with sex-neutral substitutes. If Rosser has stopped with this observation she would be making sense.
But she has gone much further than that. She would like the ETS to remove all those questions favoring boys and replace them with questions favoring girls. Her reasoning is that there is no legitimate discrepancy in test scores—girls, she points out, get better grades than boys do their first year in college, ipso facto girls’ getting lower SAT scores is clear evidence that the tests are warped.
Rosser is not asking the ETS to change those two verbal questions that girls seem to do better on, just those on which boys do better—including seemingly sex-neutral problems that boys may well be getting right simply because they have taken more math courses, or because their brains are better wired to manipulate x’s and y’s. What is she advocating but sexism, plain and simple?
Interestingly enough, the very data Rosser cites as to what questions girls and boys seem to do better on is further evidence for researchers looking for sex-related differences in interests and abilities. She also quotes ETS’s own statement to the effect that “categories designated ‘world of practical affairs’ and ‘science’ are typically easier for males, whereas the categories designated ‘aesthetics/philosophy’ and ‘human relationships’ are easier for females,” and points out, quite rightly, that any test can be skewed in favor of one group or another.
But we shouldn’t be worrying about favored groups. We should be worrying about what it is we are testing. We are not testing for virtue or wisdom or common sense, but for academic ability in certain subjects. Now, I am no fan of the ETS or of its near-monopoly on testing, and to the extent that this kind of report will clarify just what it is these all-important SAT scores have tested, I welcome criticism like Rosser’s. But when we start with the position that fairness requires girls’ scores to be just as high as boys’, we are putting the cart before the horse. (KD)