In high school and college, Adrienne Heim's academic career was stellar.

A member of the National Honor Society, she ranked among the top 10 percent of her class at her high school near Reading, Berks County. Last year, she graduated magna cum laude with a degree in education from the University of Pittsburgh. Today, she's a bilingual teacher for low-income students at P.S. 1 in the Bronx, N.Y.

But you couldn't have predicted her success if you had looked at her SAT scores.

That's because Heim, like many girls in America who take the SAT, got a low score -- just how low, she won't say. Not only that, she scored lower than many boys in her class who ranked behind her academically.

"I remember my parents being extremely disappointed and asking what was wrong, because my grades didn't match the results," she said. "I think my parents were hoping for a lot of scholarship money, maybe even at an Ivy League school. But my SAT scores pretty much ruled that out."

Heim's situation is one that some educators have recognized for years: The SAT has a gender gap problem.

Despite the fact that girls get better grades in high school and college, they are consistently outscored by boys on the SAT.

Every year, 2.3 million high school students take the SAT. According to scores released Tuesday, the average score for boys was 1,049; for girls, it was 1,006. As usual, the differences were most evident in math, where boys scored 34 points higher than girls.

	Related story Pitt revises its reporting on average SAT scores

But even in the verbal section of the test, girls -- who typically score higher on almost every other test of verbal ability -- scored nine points lower, on average.

The gap has shrunk over the years. According to the College Board, which owns the test, female test-takers' SAT math score reached a 35-year high of 503 this year.

The College Board maintains that the point spread between boys' and girls' SAT scores isn't significant.

"The SAT does underpredict for girls, but girls do better than you would expect," said Amy Schmidt, a College Board spokeswoman, noting that a score of 600 has a 30-point margin of error that goes either way, "and college admissions directors understand that."

She said the increasing diversity of students taking the test also accounts for some of the disparity. Since minorities generally score lower on the test, and more minority girls take the test than minority boys, "that's part of the reason why we have a gender gap," Schmidt said.

The College Board Web site features this statement:

"It is important to know that if groups (such as male students and female students or white students and African American students) have different average scores on the same test, it does not necessarily mean that the test is biased. If the groups actually have different knowledge and skills tied to different educational backgrounds and opportunities, the scores will reflect those differences."

But critics say such explanations only account for part of the gap.

According to a recent study of 47,000 students by the Educational Testing Service, girls taking the same advanced math classes in college and receiving the same grade as their male counterparts had scored, on average, 33 points lower than men on the SAT.

That study, said Robert Schaeffer, public education director for the nonprofit National Center for Fair and Open Testing, showed nothing new.

"Nothing has really changed" since the late 1980s, when the first allegations of gender and racial bias in the SAT surfaced, he said.

But in general, the College Board has been reluctant to act on evidence that its test is discriminatory.

Schaeffer said the board made some changes after his group filed a complaint with the U.S. Department of Education's Office for Civil Rights, charging the PSAT test -- given to 10th- and 11th-graders to determine eligibility for National Merit Scholarships -- discriminated against girls.

In response, the College Board added a multiple choice writing test on the PSAT, which helped bridge the gap by 40 percent.

Many colleges ignore the 30-point margin of error when considering SAT scores, instituting absolute cut-off scores for admission or scholarships, Schaeffer said. Despite Schmidt's contention that the differences are actually marginal, many college admissions officers tend to see those small gaps between test takers as significant -- enough to make the difference between acceptance and rejection, Schaeffer said.

The SAT -- which originally stood for Scholastic Aptitude Test but is now known by its letters only -- was developed in the 1920s as an outgrowth of the Army Alpha, the first mass-administered IQ test. The SAT was first administered in 1943 and became widely used after 1948. The test was touted as a good way to predict student performance, given the broad disparities between high schools.

But because of different learning styles, critics contend, the SAT plays to boys' strengths more than girls' strengths.

"The SAT is a fast-paced game with a premium on strategic guessing, and based on the best research we have, that better meshes with a style associated with whites and males than with girls," Schaeffer said.

Because the SAT is a timed test, girls tend to do worse because they may focus on getting the correct answer rather than risking an educated guess and moving on, maintained Alexandra Freer, author of "Girls' Guide to the SAT: Tips and Techniques for Closing the Gender Gap."

The SAT takes a "completely different approach from what you learn in school, where you work on problems carefully," Freer said. "On the SAT, you need to take a smart guess after eliminating the obvious wrong answers and going on to the next problem -- or else you'll run out of time."

Girls also tend to excel on short answer and essay questions, compared to multiple choice, "but the SAT, which is nearly all multiple choice, rewards those who take risks and guess well," she said.

Freer, who has taught SAT prep courses for the past 15 years for the Princeton Review, also noted that the SAT primarily tests vocabulary and reading comprehension, rather than grammar, writing and language use where girls outperform boys.

So what can a girl do?

Some take SAT prep classes, which can cost up to $700 or more for smaller, more intense sessions.

Carolyn Pasanek, director and owner of Mt. Lebanon Academy, which offers such classes, said she hasn't seen evidence of a gender gap. But, she added, "I think girls in some respects tend to be more emotional, and anxiety does influence test scores," she said.

And while she is the first to say her prep courses work, she admits even the relatively low $340 that she charges is generally out of reach for low-income families -- although some scholarship money is available.

"Underprivileged children are at a tremendous disadvantage for this test," said Pasanek.

The American Association of University Women -- which has long advocated for equity in education for females -- takes no official position on the SAT test. But they have been taking steps to try to boost the confidence levels of girls, including a fellowship that encourages women public school teachers to become role models -- especially in the area of math and science.

But for Heim, her SAT scores are still a source of regret.

"I never talked about it with my guy friends on the track team because I always assumed they did better than I did. I actually lied about the scores I got," she recalled.

"In one way, you do need a benchmark," she said. "You need something to rule people out. But it has to be fair, and I don't think the SAT is fair. It didn't really predict how I did, that's for sure."