Recent study exposes significant bias against applied research in computer science, a finding that spotlights the hidden hurdles women face in academia.
A new study led by Samantha Kleinberg, the Farber Chair Professor of Computer Science at Stevens Institute of Technology, uncovers a disturbing trend that may explain the persistent gender gap in computer science.
At the dawn of the computing era, women were pioneering the use of computational technology, often in roles considered secretarial. However, as computer science evolved into a prestigious field driven by algorithms and theory, women became increasingly underrepresented.
Fast forward to today, and only 23% of bachelor’s and doctoral degrees in computer science are awarded to women, with just 18% of full professors being women — an even lower representation than in the 1980s.
The study, published in IEEE Access, reveals that the type of research that often draws women into computing — applied research — is systematically devalued within the academic community.
Applied vs. Theoretical Research
In academia, research typically falls into two categories: applied research, which creates new products, technologies, or solutions to real-world problems, and theoretical research, which seeks to deepen our understanding of fundamental principles, such as the mathematical properties of algorithms.
“When you walk into a room at an applied computing conference, you’ll see a balance between women and men attendees,” Kleinberg said in a news release. “At conferences that focus more on theory, the room looks vastly different. There are significantly fewer women than men.”
Despite both types of research being crucial for the advancement of computer science, the study indicates that the academic community doesn’t value them equally.
This trend often stems from longstanding academic preferences for theoretical work requiring deep mathematical expertise, even though many researchers contribute to both areas throughout their careers.
The Bias
Kleinberg, in collaboration with Jessecae Marsh, a psychology professor at Lehigh University, surveyed tenured and tenure-track faculty across the top 100 computer science departments in the United States to gauge perceptions of researchers engaged in applied versus theoretical work.
Their findings reveal significant bias against applied researchers.
Faculty assessed researchers involved in applied work as less likely to publish in prestigious venues, secure tenure or promotion, receive awards or obtain funding.
Even more concerning, they rated these researchers as less brilliant, creative and technically skilled than their theory-focused counterparts, despite recognizing the applied work as equally important.
“I wanted to understand this dynamic I was seeing,” added Kleinberg. “So we thought, let’s find out what people actually think about this research and the people who do it.”
Analyzing data from publications, hiring, funding and awards, the researchers confirmed that applied research indeed leads to inferior career outcomes.
To test the hypothesis that women are more represented in applied research, Kleinberg manually examined over 11,000 American academics’ profiles to ensure accuracy, revealing that women are more frequently found in applied research roles, making this bias particularly detrimental to their careers.
Interestingly, universities have successfully increased women’s participation in computer science by emphasizing its applications. Initiatives like interdisciplinary CS+X programs that pair computing with fields such as anthropology, biology or music have significantly attracted more women students.
“It’s not clear whether it’s actually their interest or the culture of the field that makes theoretical work unappealing,” Kleinberg added. “It might be that women do want to do theory but feel less welcomed in those spaces.”
The study suggests that academia might be pushing women toward applied areas due to cultural barriers and then penalizing them for their work.
Broader Implications
Diverse perspectives are crucial for advancing computer science. Just as early clinical trials that excluded women led to less effective treatments for women, computing research needs varied voices to create inclusive algorithms and tools.
“I do research in health,” Kleinberg added. “Ultimately, we want our algorithms and tools to be used by everyone and to be applied to everyone. Science is better when it reflects everybody.”
The implications of this study extend beyond gender equity, warning that systematic devaluation of applied computing could deter vital research addressing society’s most pressing challenges.
Moving Forward
To address this bias, it will be necessary to make systemic changes in how universities evaluate research impact, train faculty to recognize unconscious bias, and restructure promotion and tenure decisions to value both theoretical and practical contributions.
Source: Stevens Institute of Technology