Ryan Hynd wouldn't call it an epiphany but he doesn’t deny that the research program he joined the summer of his junior year at Georgia Tech shaped his career. “It wasn't exactly a turning point, but it was like an awakening. I definitely got an idea of what it’s like to be a mathematician.” he said. “I don’t think I stopped and told myself that this was it, but that’s when I knew there was no turning back.”

Hynd, who is of Jamaican descent and now works as an assistant professor in the University of Pennsylvania’s math department, is talking about his time at the Berkeley Science Network. The program recruits underrepresented groups into science, technology, engineering, and math—or STEM—fields. Hynd was one of the lucky ones: In those areas women and African-Americans are severely underrepresented, making up just 24 percent and 3 percent of the STEM workforce, respectively.

Societal and pop-culture norms help explain why this has historically been the case. The popular ‘90s children’s cartoon Dexter’s Laboratory offers one example of the oftentimes-subtle sexist portrayals of science in pop culture. Dexter is brilliant, curious, and sharp, albeit a little eccentric. His sister “Dee Dee” (the moniker alone gives off the air of someone more simple-minded) mirrors Dexter’s eccentrism, too, but in her that characteristic is far more insidious. She doesn’t want to solve problems or even learn for that matter; instead, she spends her time ruining Dexter’s experiments. Her catchphrase, “Ooo what does this button do,” encapsulates that disposition. (Women of course suffer unique challenges in the field.) Minorities, on the other hand, rarely enjoy the ‘luxury’ of a misappropriated image in the first place; they’re hardly portrayed in the media in STEM-related fields, at least not as they are in business, music, or sports.

Attempts to mitigate this gap have made some headway. Last year, The GZA, a member of the Wu-Tang Clan hip-hop group, partnered with Columbia University to create “Science Genius BATTLES” (Bringing Attention to Transforming Teaching, Learning and Engagement in Science), a competition where students performed science-themed rhymes. However, while programs like that may work in a vacuum, on a macro scale the question of solving STEM’s diversity problem still looms.

Recent data could hold the key to closing STEM’s diversity gap in the classroom and offer insight into how different student groups learn as a whole. Kelly Hogan, a biology professor at the University of North Carolina, and Sarah L. Eddy, a postdoctoral scholar at the University of Washington, recently completed a study, “Getting Under the Hood: How and for Whom Does Increasing Course Structure Work.” The study delves into how differences in race, culture, and a family’s higher-education background can affect the methodologies by which students learn. It also encourages debate about whether college courses—specifically STEM-related ones—are using archaic teaching approaches, especially considering today’s increasingly diverse student populations.

Hogan initially became interested in the STEM achievement gap after data from own class showed that one in three African-American students earned Ds and Fs while the white and Asian children flourished. “It’s easy to think that you are a good teacher because you receive positive feedback from students but outcomes like this made me re-think if I was good,” she said.

* * *

This idea that students should not be treated as a monolithic group has been examined before: Heejung S. Kim’s 2002 study “We Talk, Therefore We Think? showed that when Asian-American students had to think out loud while solving a problem they actually performed worse than when they were allowed to work silently; white students, on the other hand, performed equally well in both situations. This revelation alone is an example of how a one-size-fits-all teaching strategy could be undermining large swaths of the student population.

To conduct the study, Hogan and Eddy decided to tweak the way Hogan taught her introductory STEM-related college classes over six semesters. Initially they kept things the same, using the traditional lecture structure as the control group: a run-of-the-mill freshmen course in a large lecture hall with limited interaction, periodic exams, and rote memorization.

They compared this with a more engaging course structure in which students participate actively in the learning process. This type of “active learning,” as its known in education circles, is proven to work, especially in STEM fields, and it differs greatly from the lecture format. Instead of dry, one-way communication—where the professor spends most of his or her time speaking at class—and grading based solely on exams, this new model lets the students mold how they learn and are assessed. These courses include preparatory and review assignments, guided reading questions, and extensive student in-class engagement.

The results of the study showed a stark difference in how changes in teaching styles affect minority students.

Average Course Grades Under Both Teaching Styles

Predicted course grades for students with an average SAT math and verbal score of 1257 (The Atlantic)

From the graph it’s clear that the new model was effective across the board, but it really worked for minorities. The gap between black students and their white and Asian counterparts (the two highest performing demographics in the class) shrunk from 5.5 percent under traditional lecture structure, to an average of 2.6 percent in the new setting.

Why do minorities improve so much relative to their peers? It could have to do with how active learning limits students’ sense of isolation and fosters communal feeling among classmates. This perhaps is where a key gap between minority students and their peers exists, as people of color and women often feel isolated. “All students viewed [this kind of] course as more of a community. Feelings of isolation weaken a student’s ability to retain information and thus develop a sense of belonging could provide the support needed to thrive,” Hogan added.

Another group that typically strives for that “sense of belonging” is first-generation students. The pressure of being the first in the family to go to college, coupled with the lack of an education culture at home, can be overbearing. Earlier this year, a New York Times piece highlighted how isolation can affect whether or not a child graduates. The problem is exacerbated when students pursue an unfamiliar field of study; students tend to resort to what they know best instead of treading the uncertain terrain of a new field.

The new course structure had an even greater impact on this group—in fact, it almost completely eliminated the achievement gap.

Average Course Grades for First and Continuing Generation Students

Predicted course grades for students with an average SAT math and verbal score of 1257 (The Atlantic)

First-generation students improved their grades by 6 percent, bringing their average scores almost completely in line with the rest of the group.

Students’ day-to-day learning improved, too, as did professors’ teaching experiences. Reading time outside of class doubled, from two hours on average per week up to five, and students were twice as likely to read assigned materials before class. And instead of spending valuable class time spoon-feeding students definitions and rudimentary concepts that they could have otherwise gained through at-home reading, professors could devote the course sessions to critical thinking—a focus Eddy sees as essential, particularly when it comes to STEM. “They need to learn to think like scientist and not like a student,” she said. “Lecturing allows students to do nothing to prepare for class and to just come in and write down the instructors ideas, instead of thinking for themselves.”

Meanwhile, traditional lecture courses are coming under fire. A study earlier this year showed that undergraduate students with traditional stand-and-deliver lectures were one-and-a-half times more likely to fail than their peers in classes that use more stimulating methods. Some experts have even gone as far as labeling lecture courses as unethical.

So why are students relegated to lectures when it’s proven that active learning can significantly enhance the educational experience? It’s worth exploring how educators learn about teaching. “Most faculty teaching STEM courses were never trained to teach inclusively and how students learn,” Hogan said. “We live in an academic culture where many faculty think, ‘This is how I was taught and it worked, so I will teach this way.’ When faculty say they should study harder they are lacking information about the diversity of students in their classroom and the knowledge about how all students learn through repeated and distributed practice.” This is an especially large problem in STEM, as the fields tend to attract pragmatic thinkers who thrived in traditional settings.

This focus on increasing diversity in STEM isn’t simply a nice thing from a social standpoint, it has practical advantages as well. Greater diversity means more innovation. Take Kelvin Doe, the self-taught 15-year-old African engineering prodigy who scoured trash bins for spare parts to build batteries, generators and transmitters and became the youngest person in history to be invited to the “Visiting Practitioner’s Program” at the Massachusetts Institute of Technology.

Back at the University of Pennsylvania, Hynd is still trying to come with a solution to the diversity problem. He was an outlier in grad school as the only black male in his class, and even now as a professor he still sees an alarming lack of women and minorities in the field. “It’s just expected that there is to be some loneliness as a minority or a woman in the STEM field,” he said.

Still, with the changing demographics of America Hynd sees a brighter future. Eventually, STEM should attract more minorities, “in fifty years it’s going to be a more diverse country,” Hynd said. “You’d only expect it. It would be a travesty if the numbers didn’t go up.”