Can Online Education Fix the Gender Diversity Problem in Coding?

The evidence of online learning in computer science suggests that success hinges less on gender identity and more on disparities in self-regulation skills. However, gender comes into play when these self-regulation skills are mediated by culture.

Dr. Amy J. Ko, Professor at the University of Washington Information School and Adjunct Professor at UW’s Paul G. Allen School of Computer Science and Engineering

Most people are aware of the gender gap in computer science and technology fields, with women just representing about one-fourth of the science, technology, engineering, and math (STEM) workforce. Many are also under the impression that the proportion of women in STEM is growing, slowly but surely.

Although the number of women receiving PhDs or certificates in STEM has increased significantly within the past decade, they still make up less than half of the number of men who received the same degrees as of 2016. More recent figures published by New York-based non-profit Catalyst reveal that men continue to dominate the STEM workforce in many countries, with women accounting for less than one-third of scientific research and development when averaged across global regions. Women were also more likely to face high rates of discrimination and bias and leave at disproportionately higher rates than their male counterparts.

In the field of computer science and coding, in particular, women are actually earning a smaller share of computing degrees than 30 years ago; the percentage of women earning computing degrees has decreased by over 10 percent during that period. And if this trend persists, the gender gap may endure for the next hundred years through to 2100, according to a paper released in 2019 by research scientists at Cornell University analyzing the decreasing parity of gender diversity in computational authorship.

We spoke with professionals in computer science and coding at two of the top ten schools teaching computer science to explore how online education can support diverse representation in these fields and promote a path of entry for women and other gender identifying people into these careers.

Dr. Amy J. Ko is a professor at the University of Washington Information School and an adjunct professor at UW’s Paul G. Allen School of Computer Science and Engineering. She directs the Code & Cognition Lab, where she studies human aspects of programming.

Dr. Ko’s earliest work included techniques for automatically answering questions about program behavior to support debugging, program understanding, and reuse. Her later work studied interactions between developers and users, and techniques for web-scale aggregation of user intent through help systems. She also co-founded AnswerDash to commercialize these ideas. Her latest work investigates effective, equitable, scalable ways for humanity to learn computing, including programming languages, APIs, programming strategies, design, and machine learning.

Dr. Ko’s work spans over 100 peer-reviewed publications, eleven receiving best paper awards and four receiving most influential paper awards. She is a member of the Association of Computing Professionals (ACM), the ACM Special Interest Group on Computer-Human Interaction (SIGCHI), the Special Interest Group on Computer Science Education (SIGCSE), and the Special Interest Group on Software Engineering (SIGSOFT). She received her PhD at the Human-Computer Interaction Institute at Carnegie Mellon University, as well as degrees in computer science and psychology with honors from Oregon State University.

Dr. Colleen Lewis is an assistant professor of computer science (CS) at the University of Illinois Urbana-Champaign. She was previously the McGregor-Girand associate professor of CS at Harvey Mudd College.

Dr. Lewis’s research seeks to identify and remove barriers to CS learning and understand and optimize CS learning. She curates CSTeachingTips.org, an NSF-sponsored project for disseminating effective CS teaching practices.
Dr. Lewis has received the NCWIT.org Undergraduate Mentoring Award and the AnitaB.org Emerging Leader Award for her efforts to broaden participation in computing. She received a PhD in science and mathematics education, an MS in computer science, and a BS in electrical engineering and computer science at the University of California, Berkeley.

“There are so many reasons,” Dr. Amy J. Ko said. “The first is simple equality; there are immense barriers to women participating in computing that have nothing to do with gender and everything to do with exclusionary cultures and harmful stereotypes perpetuated by the dominant groups of cis men that hold most of the power in computer science.”

While pursuing gender equality seems like an obvious answer, it can be reaffirming to consider the real ramifications of what persistent inequality in computer science and coding means for women.

Six out of 13 of the highest paying and in-demand jobs in the U.S. will require a degree in computer science, according to Business Insider. So, gender bias in computer science presents a real threat to future opportunity and financial stability for women.

There then are also functional and practical implications to product and system designs resulting from a lack of gender diversity in the computer science and coding workforce.

Systems of bias that discourage gender diversity have been shown to influence products and services created by male-dominated organizations, such as artificial intelligence (AI). In fields like heathcare, these implicit biases in function can endanger women’s lives. In other “best case” scenarios, these subtle skews can disadvantage women with respect to work opportunities and career growth.

“Representation in the software industry has a direct impact on what is made in the world, and how it is made,” Ko explains. “People with marginalized identities are more likely to advocate for products and services that equitably serve people with marginalized identities. If we ever hope to have a world that works for everyone, the software industry needs to represent everyone, and learn to use design methods that serve everyone, especially groups chronically ignored by dominant groups.”

Statistics clearly reflect that the number of men entering and staying in the field of coding and computer science. But what are the causes of stagnation in gender diversity in the field? Experts point to deep systemic factors as continuing to inhibit growth in gender diversity and representation in the industry.

“Cultural and structural barriers prevent women from accessing education and job opportunities in computing,” Dr. Colleen Lewis states. “As a computing community, we should do everything we can to remove such barriers.”

And this includes removing barriers and creating a culture of supporting women studying computer science in educational institutions.

Lewis pointed to a 2020 report on the status of women in computing issued by the National Center for Women in Information Technology: it found that women made up only 34 percent of those earning master’s and doctoral degrees in computer and information sciences in 2017. The study also noted that the proportion of female faculty in computing is also lower than the U.S. national average. Women faculty comprised 15 percent in computer science in 2016-2017, as compared to the 33 percent of full professors in the rest of academia during the same period.

“The data is quite clear,” Ko agreed. “Most academic CS faculty in higher education are mostly cis men. And the evidence from decades of research in gender exclusion is also clear: the culture of most CS departments is too often one that is elitist, prizes formal rigor at the expense of social relevance, and shames anyone that hasn’t had a lifetime of exposure to computers and coding. Who would want to enter and suffer such a community? Study after study documents these cultures and their impacts.”

It would also be inappropriate to fail to mention intersectionality and the overlap of gender and race bias that is highly prevalent in STEM fields, as well as the underrepresentation of people of color in those industries.

“I think that it is really important that we also focus on the cultural and structural barriers that exclude people who identify as Black/African American, Hispanic/Latinx, and Indigenous people,” said Lewis. “The underlying systems of sexism and racism are mutually reinforcing and we can’t address either one independently.”

These days, there are many online programs that can support gender diversity in learning and entering computer science and coding industries.

In the realm of online academia, Lewis highlighted some resources for changing K-12 policy and expanding access offered by the Broadening Participation in Computing, an initiative which aims to address underrepresentation in computing, including tools like the:

• BPC State Summit Toolkit: ECEP and NCWIT developed a toolkit for building a coalition for effective state education policy change to focus on broadening participation in computing.
• SCRIPT: Strategic CSforALL Planning Tool for School Districts: The CS for All Consortium has a resource for “collaborative visioning, self-assessment, and goal setting” directed at broadening participation in computing in K-12 school districts.
• State of K-12 CS Education: The Computer Science Teachers Association, Code.org, and ECEP release this annual report that describes current information related to K-12 broadening participation in computing.

There are also great examples for online programs, like the course CS Awesome, that show how the web can provide opportunities for creative, engaging, and effective pedagogical techniques, Lewis shared.

While the increased availability of online programs is useful for providing women the opportunity to learn and enter the field, the underlying systemic issues must be addressed.

“The evidence of online learning in computer science suggests that success hinges less on gender identity and more on disparities in self-regulation skills,” Ko said. “Students that can manage their learning, monitor their comprehension, and pursue resources for feedback and direct instruction, do better.

“However, gender comes into play when these self-regulation skills are mediated by culture,” Ko commented. “If students don’t perceive that a learning community is a safe place to seek help, they don’t—even when they know they need to. There’s increasing evidence that people from marginalized groups, especially women of color, often don’t feel safe seeking help in computer science learning contexts because of the risk of shaming by instructors and peers.”

Clearly, there is work to be done, specifically surrounding cultural bias perpetuated against women in computer science and coding.

“Most of the responsibility lies in schools and computer science departments to develop cultural competency for the immense diversity of people who want to learn computer science but are deterred,” Ko explained.

“The role of the government can be to incentivize this learning by faculty, as organizations like the National Science Foundation have done with mandatory broadening participation in computing plans for NSF grants in computer science. The role of not-for-profit organizations can be to create spaces for affinity groups, pathways for learners in marginalized groups to safely develop interests and mentoring relationships, and resources for helping them survive toxic learning environments.”

If these issues of systemic bias against gender diversity can be addressed, online education itself presents a promising avenue for facilitating women and other underrepresented groups’ broader representation and engagement in computer science and coding.