GENDER PLAYS A ROLE IN TEST SCORES BUT IS HIDDEN

By Anne C. Lewis for America Tomorrow

About
Anne C. Lewis





Related Web
Information:


ETS
Web Site


WASHINGTON, D.C.--When the major national standardized tests are considered, the scores of boys and girls are more alike than they are different, a fact that disproves claims that testing works against girls. However, among the "self-selected" tests such as SAT, PSAT and ACT, the ones most important to college-going students, the gender differences are more obvious.

The primary reason girls don't do as well, according to Nancy Cole, president of the Educational Testing Service, is due to the "spread." There are more boys at the bottom 10 percent and top 10 percent of the score distribution of 12th graders. The further the analysis goes on the high end, the greater the number of boys in the sample so that the ratio eventually becomes three boys to two girls. Thus, there are more males with high-performing scores than girls. Girls are closing the gap in math and science, she notes, but still have a long way to go.

Based on an analysis of 74 tests given to 12th graders nationally, Cole and fellow researchers found that females perform better in nine areas, males perform better in five areas. Girls are strongest in verbal/language skills, boys strongest in mechanical/electronic and math concepts. Girls have cut the math/science gap to only one--fourth of what it was 30 years ago, but over that time boys abilities' in verbal skills have not improved.

The gender result of open-response questions versus multiple-choice questions is mixed, but where open response calls for writing, girls tend to do better. Cole said the proposed national tests in reading and math pose a dilemma because of the study's findings. It would be desirable to have a greater proportion of the test open response (now set at 20 percent), but that would be more complicated to do. She also said that multiple-choice test results may overpredict what students can do.

The study, the most extensive yet on gender issues in testing, "confirms that both genders are failing to develop some of the desirable skills needed to succeed in our changing world," Cole told a Washington briefing.

Where there are differences, Cole attributed them partly to out-of-school interests and activities of students. The ETS study notes, for example, that females reported taking classes (music, art, dance) at higher levels for leisure activities, while boys participated in non-school sports at higher levels. Females tended to win high school awards in writing, leadership and the arts; males tended to win them in science and sports.

However, Cole also advised against using tests as the only measure of college readiness. Girls make better grades in high school, "and we have known for years that grades are one of the best predictors of college performance," she said. They indicate "studenting" skills, such as persistence and completing work. Test results, high school grades, teacher recommendations, and extracurricular activities should all be taken into account in admissions, she said. Girls also do better in college, she added.

The study, "Gender and Fair Assessment," will be published soon by the Lawrence Erlbaum Associates, Inc, of Mahway, NJ.




Return to Home Page
© 1997 America Tomorrow, Inc.
Page created May 11, 1997