Active Learning, Transformative Education

“In the first vote, maybe 50 percent get the question right, but after peer discussions, 80 to 90 percent do,” Taylor says. “It’s amazing to see the shift.” 

For Taylor, computer science isn’t just a technical skill—it’s a “fundamental literacy” for the modern world, she says. Yet the teaching of computer science is not keeping up with the rapid advance of the field.

“We have overwhelming evidence that active learning is better than standard lectures for all students,” Taylor says. “It closes race and gender gaps, helps first-generation students, and improves outcomes across the board.” 

The real challenge, she’s found, is getting professors to embrace teaching methods that work better. This conviction underpins her two-pronged research mission: to measure how much of the material students comprehend and to understand how programming concepts are most effectively taught.

Researching computer science teaching methods stemmed from her work on developing a Concept Inventory (CI) for Basic Data Structures Inventory (BDSI), which she presented at a 2019 conference. This is a multiple-choice instrument for researchers to measure conceptual understanding of core concepts—for example, linked lists, hash tables, and trees—beyond traditional test performance. Now widely adopted by entities including Google, the BDSI was developed after six years of interviews, open-ended surveys, and testing with thousands of students. Taylor published results of this research in a 2021 article in ACM Transactions on Computing Education.

More recently, Taylor’s work extends beyond assessment tools to course design. In her coauthored 2023 paper “How Do I Get People to Use My Ideas? Lessons from Successful Innovators in CS Education,” Taylor identified the “novice effect” as one of the pitfalls of computer science pedagogy. 

“Experts and novices speak different languages,” she explains. “Peers can explain things in ways novices understand because they’re closer in their learning journey. As experts, we can’t unlearn what we know, so we lose touch with what it’s like to struggle with the basics.” 

As an example, Taylor cites a 2007 Journal of Science Education and Technology article where professors, graduate students, and undergraduates watched the same lecture. The researchers were shocked to find the groups didn’t just interpret the topics differently—they disagreed about what topics were even covered.

Taylor’s research proved that interactive, student-centered teaching methods were most effective. Even small changes, like pausing during lectures to give students a moment to reflect, discuss, or think of a question, made a big difference. “It’s simple, but you’d be shocked how much more interaction you get,” Taylor says. For professors willing to go further, she recommends peer instruction, or “think-pair-share”: posing a question, then having students discuss it in small groups and share their insights with the class. 

The research also highlighted the barriers professors face when integrating new teaching methods. The optimal time to embed active learning, paired instruction, and other newer protocols is early on in course development. Departmental culture also plays a role. When one professor adopts new methods and shares materials with colleagues, professors are “more likely to try and help each other through the learning curve,” Taylor says. This peer support is especially important in the first semester, when everything feels more experimental. 

At Oberlin, Taylor has led efforts to make introductory courses more accessible by moving away from traditional math-heavy examples. “Computer science frequently has come out of math departments,” she says. “We tried to really separate out, ‘Do students need this math, or are we just using this as an example where we could use something else?’” 

Her team now emphasizes creative projects and real-world applications. Students might analyze datasets or create art through code, she says. This also helps dispel what she calls one of the biggest myths about computer science—that it’s only for “math people.” 

“People categorize themselves as, ‘Oh, I’m a humanities person; I’m not a STEM person.’ But in computer science, you’re just building a thing. It’s still a creative practice.” She would know; Taylor majored in creative writing as an Oberlin undergraduate. 

Taylor sees programming becoming as necessary a skill as writing. “So much of our lives are mediated by code,” she says. “Understanding programming, even at a basic level, makes you so much better at your job and more informed about the technology shaping our world—it’s as much art as science.”