Evidence Persists of STEM Achievement Gap for Girls
With the 40th anniversary of Title IX just days away, one key area where questions about gender equity persist is STEM education and the under-representation of women in those professions.
In researching this subject for a forthcoming EdWeek story, I discovered some evidence that a STEM achievement gap persists for girls at the K-12 level, especially in science. I also learned that what may be true for the United States is not necessarily so across the globe.
Exhibit A for the U.S. situation is recent data from the Advancement Placement program. In every STEM subject currently taught and tested, the average scores of females lag behind males, according to an analysis of data released earlier this year for the class of 2011.
That's right. Name your subject. Chemistry? Check. Biology? Check. Computer science. Statistics. Calculus. And on and on. In all 10 courses, the finding is the same: Boys on average outperform girls.
It doesn't matter if it's a discipline such as physics where far fewer girls participate, or biology, where girls are a clear majority of test-takers. In all cases, the average score for girls was lower. (Looked at another way, in every subject, a smaller proportion of girls earned what's considered a passing scoreat least a 3 on the AP grading scale, which goes from 1 to 5.)
Here's a sampling of AP score averages, drawn from data supplied to me by the College Board:
AP Physics B
AP Calculus BC
But it's not just AP data that tell this story. The same is true for NAEP, "the nation's report card," though mainly just for science.
The latest data, for 2011, show a 5-point gap for 8th graders on NAEP's 0-500 scale. (That was the only grade level tested in 2011.) Two years prior, science data for 2009 show average scores for girls trailing boys at all three grade levels tested. But what's striking here is that the gap appears to widen as students get older, from 2 points in 4th grade to 6 points by 12th.
4th grade NAEP science
Put another way, at the 12th grade, 26 percent of males scored "proficient" or above, compared with 19 percent of females in 2009.
Tracing back to 1996, the data show a fairly consistent pattern of girls trailing boys, though the size of the gap has varied. In most, but not all, cases, the difference was considered statistically significant, according to an official at the National Center for Education Statistics.
In math, meanwhile, the latest results, from 2011, show boys performing slightly better on average, with a difference of just 1 point at the 4th and 8th grades. (This difference was statistically significant, the NCES official tells me.)
But I couldn't just stop with national data. These results made me wonder about how we compare with the rest of the world. Is there a global achievement gap for females in STEM?
It depends. That's the short answer, based on where you look for answers.
On PISA, boyson averageoutperform girls in math across the 34-member nations of the Organization for Economic Cooperation and Development. In science, there is no measurable difference. (This is based on the most recent data, from 2009, for the Program for International Student Assessment, an exam for 15-year-olds.)
But guess what? Not only did the U.S. data show a gender gap in BOTH math and science (with females behind). In each case, the PISA report said, the U.S. gap was among the largest of any country tested.
In science, the biggest gaps in favor of boys were observed in the United States and Denmark. On the other hand, girls outperformed boys in five countries, including Finland, Greece, and Poland.
To make matters a little more confusing, another global database suggests that girls may have an achievement edge in both math and science when looking across nations. These data are for the most recent round, in 2007, of TIMSS, the Trends in International Mathematics and Science Study.
For math, 8th grade girls outperformed boys when looking at the average for all participating countries combined, according to an analysis by Boston College. For the United States, the gender difference was not statistically significant.) At the 4th grade, there was no measurable difference in the global average for boys and girls, but U.S. boys outperformed girls.
In science, the global average for girls was higher at both the 4th and 8th grades on TIMSS.
But the U.S. results did not fit with this pattern. At the 4th grade, the average score for boys was slightly higher but the difference was NOT considered statistically significant. At the 8th grade, U.S. boys scored higher, and the difference was statistically significant.
Here's a sampling of places with no measurable male-female achievement gap at the 4th or 8th grades on TIMSS: Hong Kong, Japan, Norway, the Russian Federation, Singapore, and Slovenia.
So, what's going on here for the United States? I took a look at several reports and consulted a few experts to help answer this question, though no one seems to have a clear and easy answer.
First off, I'll touch briefly on the question of whether there are any innate differences between males and females to consider. Certainly, the global data would cast some doubt on that notion.
I have not found an authoritative, consensus-based study on this issue per se, though I did come across a 2007 report from the National Academies, Beyond Bias and Barriers: Fulfilling the Potential of Women in Academic Science and Engineering that suggests that any such innate differences do not explain the fact that fewer women pursue advanced studies and STEM professions.
Studies of brain structure and function, of hormonal modulation of performance, of human cognitive development, and of human evolution have not found any significant biological differences between men and women in performing science and mathematics that can account for the lower representation of women in academic faculty and scientific leadership positions in these fields..
Later, the report elaborates. "Research shows that the measured cognitive and performance differences between men and women are small and in many cases nonexistent," the National Academies report says. "Furthermore, measurements of mathematics- and science-related skills are strongly affected by cultural factors, and the effects of these factors can be eliminated by appropriate mitigation strategies, such as those used to reduce the effects of 'stereotype threat.' " (I'll have more to say about stereotype threat in a moment.)
With regard to the AP data in particular, I asked Trevor Packer, the senior vice president of AP and college readiness at the College Board, for his thoughts.
First off, he said the College Board has been paying attention to the AP achievement gap for girls in STEM fields and is concerned.
"I think it's significant whenever you see different populations, ... whether different by region or gender or ethnicity, perform at different levels," he told me. "Do I have insights into exactly why that is? I have questions."
First, as to whether AP exams themselves may be biased against girls in some fashion, he said a careful vetting process by the College Board of questions ensures that this is not the case and would not explain the differences.
Second, he asked, "Is there something about the overall test-taking experience that is resulting in one group performing at a higher level, or something about the preparation that different students receive?"
And this brings up the idea of "stereotype threat," which also was tackled by the National Academies report:
Stereotype threat occurs when people feel that they might be judged in terms of a negative stereotype or that they might do something that might inadvertently confirm a stereotype of their group. When any of us find ourselves in a difficult performance situation, especially one that has time pressure involved, we might recognize that if we do poorly, others could think badly about our own individual abilities. But if you are a woman or minority-group student trying to excel in science or engineering, there is the added worry that poor performance could be taken as confirmation that group stereotypes are valid.
For a nice lay description of stereotype threat, check out this article.
Packer said the College Board is planning to launch a study this summer to see if stereotype threat might be a significant factor.
"Stereotype threat does not indicate any lack of proficiency; it indicates there is something about the experience of taking a high-stakes assessment, in which students have been inculcated with the belief that they are not adequate," he said.
Another factor Packer suggests could be at play is the increased participation of females in the AP program in STEM subjects over time.
"Anytime there is a faster growth rate in one segment [of the population], there will also be lower scores in that population," he said. "You're reaching a larger population, rather than those who persisted against all odds."
I also spoke with Martin Storksdieck, the director of the National Research Council's Board on Science Education.
"This is a bit of an unexplained phenomenon," he said of the gender gaps in STEM achievement. "My sense is that these differences are potentially a question of the kinds of tests we have, and the kinds of self-confidence people have, and expectations they have, rather than an innate difference or potentially stronger preparation that boys may have over girls."
Andresse St. Rose from the American Association of University Women, said that not all STEM data at the K-12 level point in the same direction, especially when looking at trends in coursetaking and grade point averages.
"Girls actually take slightly more math and science credits than boys do, on average," she said, "and they actually earn slightly higher grade point averages, but when you get to the AP exams, for example, you see the boys doing a little better."
That said, she still sees reason concern, and believes stereotype threat may be a key issue.
"We know that stereotypes can actually have a measurable effect on performance," she said.
Stepping back, she said: "All of us really need to examine our thinking and our beliefs and our behavior around different genders. Do we really encourage girls to do all the things that boys are encouraged to do?"
The AAUW in 2008 published a detailed report on gender equity in education, and it gives considerable attention to the STEM fields.
In closing, I'll dwell for a moment more on the international data presented here, as that seems in some ways to be especially provocative, given how U.S. outcomes compare, and contrast, with some other nations when it comes to gender.
To be sure, global comparisons are complicated. It's not just a matter of what goes on in the education system, but how that fits into the larger social and cultural context. And maybe that's exactly the point when it comes to an issue like STEM education, where many experts and advocates believe the United States still needs to see a change in attitudes, from how students view themselves to the messages they receive from educators, their parents, and society at large.
For my Title IX story, I asked Linda Rosen, the president and CEO of Change the Equation, a coalition of business leaders promoting STEM education, about changes she's seen in this area since the law for gender equity in education was enacted in 1972. Although she wasn't so sure the federal law has made much impact, she does see evidence that the United States may be turning a corner.
"I think that there's a great deal more knowledge of the need to pay attention to gender issues [in STEM education] than was the case certainly 40 years ago, or even 10 years ago," she said. Rosen, a former executive director of the National Council of Teachers of Mathematics, said she's encouraged to see more initiatives cropping up "that are really dedicated to reaching girls at a formative age, and helping them see themselves as problem solvers and design builders."
"There is a real growing understanding of this cultural stereotype that we've got," she said, "the pervasiveness of it, and the need to combat it."