Study Finds Few Signs of an Academic Gender Gap

Contradicting long-held assumptions about gender differences in academic achievement, the Educational Testing Service released a massive study Tuesday showing that there are almost no areas in which girls lag far behind boys.

The study examined the scores of 15 million students in the fourth, eighth and 12th grades on hundreds of standardized exams used by schools, as well as results from college placement exams such as the Scholastic Assessment Test.

Although boys continue to get the very highest scores on crucial college entrance tests, the study found that boys and girls are evenly matched in most important skills, including verbal reasoning, abstract reasoning, math computation and the social sciences.

“There is not a dominant picture of one gender excelling over the other and, in fact, the average performance difference across all subjects is essentially zero,” the report said.

But girls had a moderate edge in short-term memory and perceptual speed, and larger advantages in language skills, especially writing.

The superiority of girls in language ability has held up for 30 years and shows no signs of abating, according to the testing service’s findings.

In contrast, the superiority of boys in math and science was found to be surprisingly slight, “significantly smaller than 30 years ago,” the study found.

The only areas in which boys showed a clear advantage were mechanical and electronic ability and knowledge of subjects such as economics and history.

The four-year study by the Educational Testing Service gathered data from more than 400 tests and other measures and broke them down according to subsets of skills within a major subject. This produced findings that, ETS researchers said, help to debunk sweeping stereotypes about the intellectual strengths and weaknesses of women versus men, such as that women are bad at math.

In fact, the study shows that gender differences can cut both ways in a single subject.

In math, for instance, the study found that 12th grade girls perform slightly better in computation, while boys exhibited somewhat greater grasp of mathematical concepts. But in both areas, the study said, the difference was too slight to be considered significant.

Similarly, the study found that girls are not stronger than boys in all language skills. Girls were found to be stronger in writing and the use of language, while boys were better at vocabulary and “reasoning” from reading.

“To say that girls are better in some subjects or boys are better in some subjects doesn’t show the whole picture,” said Nancy S. Cole, president of the Educational Testing Service.

The study found, however, that where significant gender gaps exist, they grow over time--with girls falling most sharply behind in the most difficult subject areas that prepare students for college. By the 12th grade, the study found, more boys than girls score at the highest levels in math and science.

“It’s troubling in the sense of the options it opens and closes. That’s the issue,” Cole said.

The same pattern has been found in other large-scale studies of math and science achievement, including recent reports from the National Assessment of Educational Progress and the Third International Math and Science Study.

Disparities in achievement between boys and girls have been an increasing cause for concern among educators because of the underrepresentation of women in many professions, including law and medicine.

Accusations of test bias also have risen as educators and policymakers have placed greater emphasis on standardized testing as a means of ensuring school quality. President Clinton this year called for voluntary national exams to gauge the progress of students through high school in reading and math.

The ETS itself has been the subject of legal attacks, most recently in a case mediated by the federal Office for Civil Rights in the U.S. Department of Education. That agency announced last fall that the testing service and the College Board had agreed to add a new writing test to the PSAT--the Preliminary Scholastic Assessment Test--to eliminate bias against girls. The PSAT is used to help determine who gets hefty tuition awards from the National Merit Scholarship Corp.

But the testing service maintained Tuesday that its new study indicates that the gender differences were not the result of bias in the exams themselves. The differences, it argued, are genuine and would be reflected in well-designed standardized tests.

“Bias refers to a systematic error,” Cole said. “If gender differences are real, then a test reflecting those real differences shouldn’t be labeled as biased.”

Cole said that changing standardized exams for the sole purposed of eliminating gender differences would be a mistake because they would “hide the real differences in both directions that need to be addressed.”

It was chiefly on that point that the study drew rapid and sharp fire from critics Tuesday.

Bob Schaeffer, a spokesman for FairTest, a Cambridge-based group that strives to keep bias out of testing, called the ETS study “a smoke screen” that tries to overwhelm the public with information and deflect attention from serious flaws in its own tests. FairTest has charged that the SAT and the PSAT are heavily skewed in favor of boys, resulting in huge inequities in scholarship awards and college admissions.

The American Assn. of University Women chimed in Tuesday with similar concerns that the study would muddy the important debate over the basic fairness of “gatekeeper” tests such as the SAT.

“While it is reassuring to learn that in the great universe of standardized tests the [ETS finds] little gender differences in overall test performance, as a high school principal I am still troubled by the fact that boys score roughly 40 points higher than girls on the PSAT and SAT, which lead to an inequitable distribution of scholarship dollars,” said Jackie DeFazio, president of the association of women college graduates.

The AAUW conducted its own study five years ago, culling two decades of research on gender bias, and concluded that most standardized tests discriminate against girls. In an earlier poll it found that girls, unlike boys, generally finish schools with less confidence and self-esteem than when they began their education.

Times staff writer Heather Knight in Washington contributed to this report.