688
PARTISAN REVIEW
duced disparate outcomes for different groups was
ipso jacto
biased . Even
when the bias was not apparent, the very fact of different outcomes was
proof of bias.
Since the tests, it was claimed, were unfair to minorities, those in
pursuit of not just equal educational opportun iti es, but equal educational
outcomes for all groups proposed two complementary tacks. One was
affirmative action in admissions and hiring; the other was test-norming
for different groups . Both would produce the exactly equal results for all
groups that never had been produced by any form of objective testing.
It
is worth looking closer at the issue of objectivity. One of the
tenets of political correctness is that objectivity is impmsible, and that
any system or selection process or test will always be biased
to
favor
those who are in power. Characteristic of political correctness is the as–
sumption that any difference in performance demonstrates bias in the
measure. By now, we have heard so many attacks on objective measure–
ments that it is no longer novel when claims of bias are raised. Given the
fact that there has been real discrimination in our history, it is reasonable
to suspect lingering bias whenever charges are lodged in the press.
Whenever the pass rates for groups differ, we can expect to hear claims
of test bias, regardless of the test.
Take the issue of gender bias. It is well known that boys get higher
scores than girls on tests of mathematics and science; it is not so well
known that girls get higher scores on tests of reading and writing.
Numerous studies contend that the SAT is biased against girls because
their scores are not as high as those of boys; the gap is greater in mathe–
matics and relatively small on the verbal section (where girls' scores were
higher than boys' until 1972) . Researchers claim that the content of the
math questions favors boys. They say that references to sports, the mili–
tary , or finance give unfair advantage
to
boys, while girls do better on
questions that refer
to
cooking, clothing, and jewelry. It should be em–
barrassing to researchers to reinforce gender stereotypes. Why shouldn't
girls be expected to figure out the win-loss record of a basketball team
(it might even be a
girls'
basketball team)? How will women succeed in
today's workplace if they can solve only problems that are connected to
the kitchen or their wardrobe? Ironi cally, many of the math questions
on which girls do poorly have no words at all, but are number prob–
lems . According to the researchers, the problems are biased anyway be–
cause girls can't solve them as often as boys. Even if all the word ques–
tions were eliminated, boys would still get higher scores than girls. I do
not know why this is so.
It
may have to do with cultural attitudes that
are shared by students, parents, teachers, and the larger society about
what kinds of school subjects are for boys and what kinds are for girls. It
