Tests' Sex Bias Costs Women Millions In College Assistance
Young women lose millions of dollars in college scholarships each year as a result of sex bias in admissions tests that are widely used in making such awards, argues a report released last week.
"These tests are supposed to predict future academic performance,'' said Phyllis Rosser, the principal author of the study by the National Center for Fair & Open Testing, or FairTest. "Yet, girls earn better grades than boys in both high school and college, but score lower on standardized tests.''
"The more these biased exams are used as gatekeepers for college admission and scholarship aid,'' she contended, "the more discrimination young women suffer.''
For example, the report says, of the approximately 15,500 semifinalists competing for National Merit Scholarships this year, 61 percent are men and 35 percent are women. More men than women are semifinalists in every state except Hawaii and Montana, according to FairTest's calculations.
National Merit semifinalists are chosen on the basis of their scores on the Preliminary Scholastic Aptitude Test. Because women generally score lower than men on the P.S.A.T., they are less likely to become contenders for the prestigious awards, FairTest says.
The advocacy group identified the sex of the semifinalists from a list of their names. It was unable to determine the sex of approximately 4.3 percent of the competitors.
The group also reports that in 1985-86, girls received only 36 percent of the National Merit Scholarships, while men received 64 percent. That year, it notes, some 643,000 women took the P.S.A.T., compared with 537,000 men. The sex of approximately 1,000 additional test-takers was not identified.
In addition to the National Merit Scholarship Corporation, more than 800 other private and government scholarship agencies rely at least partially on admissions tests to select winners, according to the study, "Sex Bias in College-Admissions Tests: Why Women Lose Out.''
Scholarship Group Responds
FairTest's charges drew a quick response from Marianne Roderick, senior vice president for the nonprofit organization that provides more than $23 million in National Merit awards to students each year. Some 670 corporations, foundations, professional organizations, colleges, and universities sponsor the scholarships.
Ms. Roderick stressed that, while P.S.A.T. scores are used to select semifinalists, the organization also considers a wide range of other criteria in making the awards, including students' academic records, extracurricular activities, intended college majors, leadership potential, and recommendations of their school principals.
"Because the test scores of all finalists in the merit program are very similar and very high,'' she said, "the other factors really become more important in choosing the scholarship winners.''
Ms. Roderick said the P.S.A.T. is used as an initial screening device because it is readily available to all interested students and allows them to enter the competition on the same basis.
But she admitted that in the past few years the annual pool of approximately 15,000 semifinalists has been about 60 percent male.
Ms. Roderick declined to address the charge that the P.S.A.T. is biased against female students. That question, she said, should be asked of the Educational Testing Service, which produces the test.
"We are not in the testing business,'' she said. "What we operate is a scholarship competition.''
Fred Moreno, assistant director of public affairs for the College Board, which contracts with the E.T.S. to produce both the P.S.A.T. and the Scholastic Aptitude Test, said, "Fundamentally, we don't think the [P.S.A.T.] is biased.''
There is a difference in how men and women score on both the P.S.A.T. and the S.A.T., he said, "but we don't know all the answers and reasons why.''
For example, he noted, "at least in terms of the math section, we know that women tend to take less math and science courses in high school than men do, and that could have an effect on their math scores.''
Mr. Moreno added: "There are some 1.2 million P.S.A.T. takers who are applicants for National Merit Scholarships. It would be impossible for [the National Merit Scholarship Corporation] to look at the portfolios of 1.2 million people.''
He said an analysis by the E.T.S. had found that using the P.S.A.T. as an initial screen for the scholarships did not appear to be hurting students. Scholarship winners, he noted, are judged on a wide variety of measures, "and boys still wind up getting more scholarships'' than do girls.
According to FairTest, the proportion of female National Merit scholars has decreased in the past three years. In 1985-86, of the 6,026 scholarships awarded, 36 percent went to women, compared with 38 percent in 1984-85 and 40 percent in 1983-84.
Its report includes a state-by-state breakdown of this year's National Merit semifinalists by sex; a list of corporations that give special merit scholarships administered by the National Merit Scholarship Corporation; and a list of other scholarships awarded, at least in part, on the basis of college-admissions-test scores.
The FairTest study also maintains that widespread use of college-admissions tests--including the S.A.T. and the American College Testing program examination--denies women access to competitive universities and college-preparatory programs.
According to the report, more than 1,000 colleges and universities require that scores on college-admissions tests be submitted before they will consider an applicant.
FairTest lists nearly 300 accredited colleges and universities that it says rely heavily on test scores in making admissions decisions.
In addition, FairTest reports that a number of special high-school programs for the "gifted and talented'' use college-admissions scores to select participants.