Gender Gap
COLLEGE TEST BIAS CONTINUES UNABATED
Why does the gender gap in SAT scores persist?
By Andrea Coombes, CBS MarketWatch.com
Last Update: 12:01 AM ET Sept. 15, 2003
SAN FRANCISCO (CBS.MW) – For 32 years, boys have outperformed girls on the verbal portion of the college-entrance exam known as the SAT, and boys' dominance on the math portion has been pronounced for at least 36 years.
Yet, while the SAT claims to be a predictor of college grades -- and is used as such by most colleges and universities nationwide -- girls have consistently outperformed boys in college classes for years.
About 12,900 young women did not get into top colleges in the late 1990s due to the test's gender gap, according to one U.C. Berkeley study. The problem occurred mainly at large, highly competitive public schools facing large numbers of applicants -- and thus more likely to use SAT scores as an initial sorter, the study found.
While this year's overall test results are improved, the gender gap widened slightly this year over last year. Girls' average verbal score this year is 503, compared with boys' 512. That 9-point difference is up from a 4-point spread last year.
The 34-point spread between girls' and boys' math scores this year matched last year's, with both groups' scores rising 3 points to 503 for the girls and 537 for the boys.
The test's purveyors don't deny the existence of the gap. In college, "girls come in better than you would expect and boys slightly worse than you would expect" given their SAT scores, said Amy Schmidt, executive director of higher-education research at The College Board, the nonprofit association that owns and administers the test.
But the question remains: Why has the gap persisted for so long? The answers to that question are hotly contested.
"We don't see this as a technical flaw of the test," Schmidt said. Instead, she said, a growing pool of female test-takers, one that's more diverse and includes more low-income students than the boys group, is pulling down girls' scores.
"Are we doing anything about it? No. We don't believe it's a flaw of the test," she said.
Others contend the gap has occurred for too long for the growing number of female test-takers now to be relevant. "Each decade they've had a different explanation for the gap," said Alexandra Freer, author of "Girls' Guide to the SAT" (Random House/Princeton Review Books, April 2003).
"Before it was, 'Girls don't take as rigorous math courses.' No matter what their explanation, the gap hasn't gone away," she said.
Instead, Freer said, the test itself is largely the crux of the problem, because it gives higher points for the type of test-taking abilities that boys tend to demonstrate.
For instance, the test is "speeded," which means those who are more comfortable guessing will often do better by completing more problems.
High-school girls interviewed by Freer said that even after narrowing a problem down to two possible answers, "they would continue to try to solve the problem, versus taking a guess and moving on," she said.
With girls outperforming boys in high school, "clearly that type of test-taking skill and learning behavior is rewarded throughout high school and college. But it's not rewarded on the SAT. In fact, it can penalize your score."
Change in the works
In 2005, the SAT will contain a new writing portion and by all accounts this should help raise girls' scores, experts said.
The College Board was required, as the result of a lawsuit in the mid-1990s, to include a written portion on the Preliminary SAT, a test given to high-school juniors and used to assess eligibility for National Merit Scholarships.
That ruling was based on the finding that scholarship recipients were mainly boys as a result of the gender gap inherent in the test's scores, said Bob Schaeffer, public education director at the National Center for Fair and Open Testing, a watchdog group that monitors testing programs nationwide and helped bring the lawsuit.
The PSAT comprises old SAT questions.
The PSAT's new writing portion "closed the gender gap on the PSAT by about 40 percent, which results in several million dollars a year of additional scholarships going to females," Schaeffer said.
But the new written portion on the SAT is not being added because of the gender gap, the College Board's Schmidt said. Instead, "we feel strongly it's an important component of college success skills."
While the writing portion will lessen the gender gap, it likely won't eliminate it. "That's an improvement, a big improvement, but it's probably not the whole thing," said David Leonard, a political-science professor at U.C. Berkeley and author of the study assessing the impacts of the SAT gender gap.
He says no single change would cure the test of its gender-gap bias. After conducting the study, "it became evident to me that it was so complicated to try to fix the test that I wanted to get more publicity to the existence of the problem, let universities know how to handle it. What was distressful to me was how few admissions experts were aware of this problem."
The test administrators, said Leonard, "at a minimum need to at least publicize the existence of this gap more vigorously."
Overall, boys are faring worse
To some, the SAT gender gap is a moot point. Girls are attending college and earning degrees at increasingly higher rates than boys, they say.
Women earned about 56 percent of bachelor's degrees and almost 58 percent of master's degrees in 2000, according to the National Center for Education Statistics.
Still, others say the gap in SAT scores represents a separate issue. "There's a crisis with boys in education. That's absolutely true," Freer said. But, while women are clearly getting into schools, their SAT scores may be affecting which ones.
"Are they getting into the schools that they are qualified for? Are they getting the scholarship money? Are they getting into the right programs?" Freer said.
It's not only school admission and scholarship money at stake, some point out. A girl taking the test could face a 100-point difference compared with a boy on the combined math and verbal portions of the test, Leonard said.
Put yourself in her shoes, he said. "That means that maybe you and a boy in your high school class who were doing exactly the same in high school, he ended up with 1400 and you ended up with 1300. Both are great scores, but you probably felt you weren't as smart as he is."
Andrea Coombes is a reporter for CBS.MarketWatch.com in San Francisco.