Girls seek to narrow gender gap on SAT

By Laura Pappano, Globe Correspondent, 10/5/2003

Education has its fashions, but some things we have come to take for granted: Kids work together on homework and it's not cheating. Teachers have e-mail. Whiteboards are as common as chalkboards. . . . And girls score lower than boys on the SAT.

This last fact -- the perennial point gap between male and female scores that has persisted for decades -- was once cause for public outcry. But lately, it has been a nonstarter.

"What's happened is women are now 56 percent of those receiving bachelor's degrees, and nobody cares that they're not doing as well on the SAT," said Phyllis Rosser, a scholar and director of the Equality in Testing Project in New York who has tracked the issue since 1979.

In other words, public concern has moved on. Never mind that girls arriving with sharpened pencils to take the Scholastic Assessment Test on Saturday will take a test in which girls historically score lower than boys by 40 points -- even though research shows that girls earn higher grades in school.

Scores for the class of 2003, released in July, showed boys in Massachusetts earning an average 522 verbal score and 539 on math. Girls scored 511 on verbal and 507 on math -- an 11-point verbal gap and 32-point math gap, which is similar nationally. (Each test has an 800-point scale.)

The gap has become so widely acknowledged that one test-preparation company, the Princeton Review, recently published "The Girls' Guide to the SAT," to arm girls with tools to counter their disadvantage.

Author Alexandra Freer, a veteran test-prep teacher who interviewed 150 girls and surveyed 300 high school students for the book, believes elements of the test favor males, including the need for speed and effective guessing. But Freer also believes girls are socialized in ways that reward them in school -- and hurt them on the SAT.

"The SAT is not testing that you can do a problem, but that you can do it quickly," said Freer. "One of the reasons girls do well in school is they tend to be more meticulous and detailed. . . . They solve their problems completely, show all their work. Exactly the opposite is rewarded on the SAT."

Instead of approaching a math equation by solving it and then comparing the answer to multiple-choice answers, she said, it is wiser to plug in answers from the list given to find which fits -- something girls typically don't do.

Another problem many girls face is a lack of confidence, despite strong academic skills, said Lisa Jacobson, CEO of Inspirica, a tutoring and test-prep company in Manhattan and Newton. Girls the company tutors "come in thinking the test is going to beat them," she said, while boys "come in thinking they are going to beat the test."

Amy Schmidt, executive director of higher education research at the College Board, which creates the SAT, said score differences between the sexes exist because the pool of females taking the test is larger and more racially and socioeconomically diverse than the pool of males. Girls represent 54 percent of the 1.2 million students taking the SAT each year.

"When we create the test questions, they go through an enormously complex process," said Schmidt, who said adding a new writing section in 2005 that includes a 25-minute essay should narrow the gender gap. "The items are prescreened to ensure there is no bias."

Nonetheless, researchers find a key contradiction: The SATs role as a predictor of freshman-year grades runs counter to the fact that girls, even while scoring lower than boys on the SAT, earn higher first-year college grades.

And girls do not fare worse than boys on many other tests. Results of the National Assessment of Educational Progress, for example, show a statistical dead heat between 12th-grade boys and girls on the math test. The 2003 MCAS yields similar results: 51 percent of 10th-grade girls and 52 percent of boys scored advanced or proficient in math. In English, 67 percent of girls and 56 percent of boys scored advanced or proficient -- a difference that makes the lag in verbal SAT scores more puzzling.

Schmidt said that a selective group takes the SAT while tests like the NAEP reflect a broader population. But researchers who study the issue conclude that the SAT underpredicts girls' college grades. The debate over whether this matters has faded because women -- since 1979 -- outnumber men on campus. But Freer said that unfairly lower SAT scores prevent women from getting into the colleges they deserve.

That notion is supported by David K. Leonard, political science professor at the University of California at Berkeley, and associate Jiming Jiang, who studied SAT scores and college admissions. They say that the tendency of SAT scores to underestimate girls' GPAs results in 12,900 females a year not admitted to the student bodies of large public colleges, "with 10,600 missing at the very competitive ones."

Freer and others also say that lower-than-deserved SAT scores prevent girls from gaining scholarships, financial aid, and positions on college sports teams for which minimum SAT scores are required for participation. Lower scores also have the subtler effect of spurring girls to downgrade their aspirations.

"There is a huge negative psychological feedback in which young women who have done very well in high school get their test scores and lower their sights because they internalize their scores and believe them," said Bob Schaeffer, public education director of FairTest in Cambridge, which this week releases an updated list of nearly 700 colleges that do not demand SAT scores for admission.

Public attention to the SAT gender gap may have faded, but students like Wayland High School senior Ariel Tichnor are paying attention. On Saturday, she is taking the SAT -- again. Tichnor's grades put her in the top 5 percent of her class. But her first round of SAT scores "don't reflect how well I do in school, how hard I work," she said.

Wayland High's Guidance Director Norma Greenberg has not thought much about the gap, and was surprised when she pulled up the school's numbers.

The mean scores for Wayland High's class of 2003 show girls scored 27 points lower than boys in verbal and 44 points lower in math. The gap for the class of 2002 is 27 points in verbal and a whopping 53 points in math. Yet Greenberg said those classes were not dominated academically by boys, and the top 10 percent were split about equally between male and female students.

The SAT gender gap is a problem because it places a group of students who generally perform better than scores suggest at a disadvantage.

Rosser, of the Equality in Testing Project, said test makers know which questions favor which sex. "They can even the score in a minute, and they have chosen not to," she said.

How much of the gap is the result of how girls are taught or socialized? Kenneth J. Meier, professor of political science at Texas A&M University, said his study of math performance, including on the SAT, showed girls earned higher scores when they had female teachers. Whatever the cause, colleges that use the SAT for admissions should acknowledge and make allowances for the persistent gender gap.

With schools like the University of Pennsylvania, Washington University, and Northwestern on her list, Tichnor needs her SAT scores to reflect -- not camouflage -- her academic talent.

Send feedback and ideas to chalkboard@globe.com.