Women learn less than men in college - about one-third less, according to new research peering into the mysterious realm of what students actually learn on campus.
Discovering just how much undergraduates learn in English, math, science, and social studies was the aim of a study of 19,000 students at 56 four-year colleges and universities in 13 states.
But when researchers compared students' scores on a standardized test, one finding leaped out: gender was a huge factor in how much those scores improved over time. Women's scores improved only two-thirds as much as men's over the course of four years. Women lagged most in math and science, but also in other areas.
"To me, the finding is disconcerting," says Ernest Pascarella, a University of Iowa professor of education and co-author of the study.
"We're the first to have found this gender effect, at least as far as I know," he says. "I'm still a tad skeptical until someone else has similar results or we do another study and find the same thing."
Others are skeptical, too. The standardized test could be gender-biased, they suggest, or the different scores could be explained at least in part by the fact that women take fewer science and math courses than men.
"These results are from one test," says Valdrie Walker, an associate professor of education at Sweet Briar College, an all-women's school in Virginia. "What are the qualitative variables?"
Study findings were reported in the peer-reviewed Journal of Higher Education. And researchers say the test they used - the College Basic Academic Subjects Examination - has been vetted by scores of scientists over a number of years. But they do offer caveats.
The student sample was not random - even though the race and gender breakdown had "reasonable comparability" with the national profile of traditional-age college students.
They also acknowledge that the data did not indicate students' courses and majors. Still, Dr. Pascarella and colleagues stand by their analysis, even as they call for further research to verify or dispute their findings.
"The study is not saying that men are somehow smarter than women," Pascarella says. "We don't know what these results mean. But we do know that lots of times, our [American higher education] pedagogy is male oriented, and that may be part of what's happening here."
This latest study is only a small part of a shift in education research toward the critical question: How do you measure what students actually learn in their undergraduate years?
If successful, such efforts could yield college rankings based on how well schools teach their undergraduates - not just the number of books in the library or the size of an endowment.
Much of the unhappiness over rankings today is because higher education is "a black box" - and rankings like those by US News & World Report only hint at what's inside, researchers say.
"People assume that if a young person goes to Harvard, Wellesley, or Swarthmore they will be better off than at a state public university," says Roger Benjamin, president of the Rand Council for Aid to Education in New York. "What the institution adds to the individual student is never really answered."
But can research really do that? At least a half-dozen well-funded organizations believe they can, and are working on tools to assess reading, writing, and critical thinking undergraduates in an objective and quantifiable fashion.
At the Educational Testing Service in Princeton, N.J., researchers have made key "breakthroughs" in developing machine-scorable writing tests at the higher-education level, observers say.
One of the most advanced efforts is the National Survey of Student Engagement (NSSE) funded by the Pew Charitable trust, which released its second set of results last week on student performances at 470 four-year colleges and universities.
The NSSE isn't a test of outcomes, strictly speaking, like Pascarella's work. Rather, it's a measure of student perceptions. Its survey asks about time spent on activities - homework, reading, writing - that have proved to have an impact on learning.
"It may turn out that women do take less math and science - and there may be a system bias in many institutions, or a cultural climate where women are discouraged from continuing in math or science," says George Kuh, director of the NSSE.
Marcia Baxter-Magolda, a professor of education at Miami University in Oxford, Ohio, is also probing gender issues in higher education. She has published research showing gender-based differences in "ways of knowing" - internal assumptions she says affect how people learn.
Most college classrooms, she says, are still structured in ways that play to the advantage of what she calls "separate knowers" - those who instinctively disagree, challenge, question in class. This group tends to include more men than women, she found.
Many college classrooms don't, however, favor "connected knowers," which she describes as those who thrive on learning by working through problems in groups. More women than men pattern this way of knowing, she says.
While more research is needed to validate the recent study by Pascarella and his colleagues, Dr. Baxter-Magolda says it would not surprise her if there did turn out to be problems with classrooms being structured in a way that favors men.
"Many have written about girls in secondary schools who are shortchanged in their educational experience," Baxter-Magolda says. "This [new] research suggests that problem may continue in higher education. I'm not sure that even men are getting the best educational experience they could."
The idea that a woman might get a better education by shutting the classroom door on men is hardly new.
For decades, it's been the subtext as women's colleges woo prospective students away from the much-larger co-ed universe.
"The advantages of women's colleges are hard to match in the co-ed world," opines the website for the Women's College Coalition, a Washington-based group representing 70 women's colleges.
Even so, Jadwiga Sebrechts, president of the coalition, is skeptical of claims that women learn in different ways from men. She does say women learn better in a single-sex environment - but mainly because "it values them more."
"When women are in an environment that really validates and affirms them as thinkers and potential experts," she says, "[they are] more engaged, more motivated - and, in the end ... learn more."
The same type of research that recently compared how much women and men learn at co-ed colleges could be used to measure if women really learn more in a single-sex institution, observers say.
In certain fields, it's clear that women's colleges yield different results. Women there are three times more likely to earn a B.A. in economics and one and a half times more likely to earn a degree in life sciences, the physical sciences, and math than women at co-ed institutions, according to the Women's College Coalition.
Women also have more role models at women's colleges, many say.
Women make up half the mathematics faculty at Mt. Holyoke College, a women's school in South Hadley, Mass., compared with none at all at many schools, says Donal O'Shea, dean of faculty.
If women learn less than men in co-ed colleges, as a recent study suggests, Dr. O'Shea says one factor may be the loud- mouth undergraduate male. "Women are more cautious about what they say.... If you put a 50-50 group of men and women together, the men do more of the talking because they don't mind looking like idiots at times." Women's voices tend to get lost.
There's no need to convince Kathleen Homan. The senior at Mount Holyoke College picked the school because, after watching the interaction between professors and students, she decided it would be a better learning environment than any of the four co-ed schools she was considering.
Now, her reaction to the recent study is that maybe her gut feeling was right on target. "I feel I've learned a lot more [here] because it is a women's institution," she says.