Can Online Education Fix the Gender Diversity Problem in STEM?

With online education, we can scale the best female instructors in diverse subject areas to introduce more women and learners in general to instructors like them, providing high quality role models.

Vinod Bakthavachalam, Senior Data Scientist at Coursera

According to the National Center for Education Statistics, women earned roughly half of all undergraduate science and engineering degrees in 2016. In those same fields, they earned 44 percent of master’s degrees and 41 percent of doctorate degrees. The NCES found that female STEM students typically pursue psychology or the biosciences. It also found that women’s lowest STEM enrollment figures are in the areas of engineering and computer science. But what could be the source of this imbalance?

Although the gender diversity problem in STEM has been well-known for some time, it’s only recently that researchers have been able to pull back and see the big picture. Using its own enrollment data, the online education provider Coursera has studied how this issue affects the American economy. It found that having female instructors helps close the gender gap in STEM fields. Generally a male-dominated network of disciplines, STEM has finally begun the long upward trajectory toward gender parity.

Read on to learn more about how online education can help shrink the gender gap in STEM studies.

Vinod Bakthavachalam, Senior Data Scientist at Coursera

Vinod Bakthavachalam has worked for Coursera for over four years years as a data scientist. Before that, he worked in quantitative finance and studied economics, statistics, and biology at the University of California, Berkeley and Stanford University. During his time at Coursera, he has examined ways to close the gender gap in STEM and increase equality in online learning.

A number of nations have shown that STEM gender equality is both possible and preferable. However, the NCES figures essentially match the male-to-female occupational STEM disparity reached by the World Economic Forum.

As a student, Vinod Bakthavachalam noticed an imbalance: “I had very few female instructors in my statistics and economics classes, but more in biology,” said Bakthavachalam. “I think this is a common experience but not something you think about much of the time. It is like an unconscious bias and something that our experiments, and much other research, has shown to be important.”

It’s this kind of unconscious bias than can it make so challenging for women to pursue STEM studies. Bakthavachalam learned about how female instructors, serving as role models, can help to close the gender gap. “Our work in this area has focused on identifying reasons for the lack of gender equality in STEM and ways to close the gender gap,” Bakthavachalam said.

By following trends in online education via cutting-edge platforms, STEM industries can glean valuable insights about the changing professional landscape: “Surveys and observational data can be helpful,” Bakthavachalam stated. “Something as simple as looking at the funnel from enrollment to completion can shed light on the places to focus resources.”

But one of the most significant discoveries in Bakthavachalam’s analytical work concerns the mentors, teachers, and leaders that young women need. Like matters of cultural representation in the media, young people need to be able to learn in a field that actually acknowledges female contributions. And they need to learn, in part, from individuals who share their gendered experience of the world.

“One big finding in our work at Coursera has been the power of role models in explaining part of the gender gap in STEM,” shared the data scientist. “Female instructors cause women to be more likely to enroll in STEM courses, though many universities lack female faculty in STEM areas. With online education we can scale the best female instructors in diverse subject areas to introduce more women and learners in general to instructors like them, providing high quality role models.”

In what ways can online studies in science, technology, engineering, and mathematics help to equalize the gender diversity gap? Bakthavachalam and his team spearheaded a number of studies addressing this question.

“For online research, there is the benefit of being able to run structured experiments that can measure the causal impact of various interventions and identify what policies work versus don’t. In addition, the scale of online research (large sample sizes) unlocks the kinds of questions you can study.”

So, with access to big data easier than ever, data scientists like those at Coursera can help contextualize how gender functions in the sciences. One example of Bakathavachalam’s work toward a solution involves analytics: “Generally, gender diversity on Coursera is on par with gender diversity on campus,” said Bakthavachalam. “Fewer women end up enrolling in STEM courses compared to men on Coursera.” This tends to follow the same pattern as on-campus STEM enrollment.

“However, the share of female instructors in STEM topics on Coursera tends to be higher than on campus,” added Bakthavachalam. “This in combination with the ability to personalize the learning experience allows us to build a product that encourages women to enroll and persist in STEM content. We have piloted initiatives to do just that.”

In their studies, Bakthavachalam and his team have identified five factors that are critical to closing the gender gap in STEM studies. They are:

1. Scale female instructors.
2. Design course material for inclusiveness.
3. Use machine-learned interventions to drive persistence.
4. Launch new features to support a diversity of learning strategies.
5. Enable university experimentation and iteration.

Taking these steps gives female students opportunities to connect with people who share part of their life experience. Varying economic and social modes influence young women’s interest in the sciences. If they see women as instructors and in positions of expertise or authority, it can help to foster a more gender-inclusive environment.

As mentioned above, Coursera’s study on female role models in STEM fields shows a need for more equal gender representation.

“For learning online, there are benefits of personalized learning to ensure struggling students can get personalized help and coaching, better recommendations for learning targeted to their background and interest,” revealed Bakthavachalam. “Other benefits are online learning’s lower cost, the derisking of the learning process, the fact that it is anonymous, and the ability to better scale role models.”

Going forward, there is no doubt that women’s representation in the sciences will continue on its trajectory toward greater equality. The important work done by Coursera on both the educational and analytical fronts has helped to pave the way for more widespread discussion of this issue. Bakthavachalam is excited and optimistic about what the future holds.

“I would say that data science can help address diversity because it has a toolkit that can enable us to measure and track a lack of diversity, and design experiments and products to combat it, but it is not something data can do by itself,” said Bakthavachalam. “There is lots of cross functional work needed to ensure the findings in data make it into a company’s product and marketing, and much more we need to do in the education system broadly.”

Those broad, systemic changes can be affected by a number of initiatives working together, and equal representation in the field is the place to begin. The toolkit that Bakthavachalam mentions gives institutions, educators, and researchers a complex glimpse into the factors influencing female STEM enrollment.

As online education platforms continue to gain traction with young people, the 21st century is set to observe a number of interesting developments to training and skilling.

“Data science and innovation in general thrives on the ability to make sense of complexity and solve problems,” Bakthavachalam said. “Still, despite the complexity, we need to ensure we are increasing diversity. To do this effectively, we need diverse perspectives that can see things from different angles and ensure everyone’s perspective in represented.”

Organizations like Coursera offer both a curriculum and solutions to today’s pressing educational problems. On his blog, Bakthavachalam writes about using technology to increase equality. In conversation, he noted that his “work has shown that the gap starts even before enrollment when people are mentioning what subjects they are interested in.”

This seems to suggest that online education alone can’t completely solve the matter, but Bakthavachalam qualified those future solutions: “There are also things we can do online to close the gap because the funnel does widen from enrollment to completion.” While some of these approaches involve business, others involve the educational choices of students while they are very young.

Overall, by creating more inclusive academic and professional environments, we can encourage an increase in gender diversity across a wide number of STEM industries.