How three STEM lecturers turned traditional lecturing into active learning moments
In this best practice story, three Science Faculty lecturers decided to radically “activate” their lectures: Ellen den Ouden (BSc Mathematics), Lise Stork (BSc Information Science), and Krystal Guo (BSc and MSc Mathematics). They kept the lecture format as a base, but redesigned what happens in those two hours so that students think and work almost continuously.
They alternated explanations with short or longer tasks for students, to immediately apply their knowledge and develop their thinking and application skills. Whether using state of the art AI, relevant models or visuals, or old-fashioned paper and pen, each lecturer got the students to actively work with the content in their own way.
Courses: Ellen den Ouden – Vectorcalculus, BSc Mathematics, year 2; Lise Stork – Semantic Web, BSc Information Science, year 1+2; Krystal Guo – Combinatorial Enumeration, BSc Mathematics, elective + Graph Symmetries and Combinatorial Designs, MSc Mathematics year 1
Active learning designs: Three ways of getting student brains to participate
Ellen and Lise independently developed a similar set-up, in which they alternate short blocks of theory lecturing with activities that students do alone or in duos. And while Ellen and Lise use short passive-active cycles, Krystal chose to leave even more of the intellectual work in her lectures, which consists of building proofs, to the students themselves.
 Ellen den Ouden |
 Lise Stork |
 Krystal Guo |
Active design 1: From continuous talking to active-passive learning loops
Ellen teaches a calculation‑heavy mathematics course, often at 9:00 in the morning. “This is not generally a moment when students come to class very excited,” she notes. Still, her attendance hovers around 60%, and students sometimes come only for her 9–11 a.m. lecture block. “I do see that as a compliment to my teaching.”
Her lecture style consists of mini‑explanations followed immediately by work. “I explain the concepts only really briefly; 5 to 10 minutes max. Then the students work for 5 to 10 minutes too.” It creates a balanced division of time. In her slides, Ellen puts short questions directly connected to the concept she just explained. After students work on those individually (on paper!) she discusses the answers with the group. That way, students have practiced the basics during the lecture so they can go deeper in tutorials.
Lise, teaching Semantic Web, developed a similar set-up. “I have lectures where I explain a theory, and after that, students pair up and do a task. It didn’t seem correct that I was the only one talking during a lecture; that didn’t give me any idea whether students were engaged and actually learning, so I wanted to get the students talking too.”
She structures her lectures in blocks. After each piece of content, students get about ten minutes to work in pairs on laptops. Sometimes they download something or work with a model or a visualization; the specific instructions are in the slides. When time is up, she asks a few pairs to share their answers, and other students react or add to them until they reach an answer together.
This structure also made Lise rethink the content. One particularly dense lecture originally contained many comparable examples of reasoning axioms, “They are actually very similar,” she realised, “yet they were all taught individually one after another!” She used AI behind the scenes to cluster these axioms into a few intuitive groups, then focused her lecture on those core ideas. Students chose one of them that interested them, practiced in pairs, and created their own real‑world example. Lise walked around, asked some pairs to share, and facilitated discussion on how to improve or refine their examples.
Active design 2: A theorem as a ‘choose your own adventure’
The course material in Krystal’s BSc course material is not computationally heavy, but very theoretical. “Students find this difficult,” she says, “because they have to argue using words rather than doing algebraic manipulations to equations.” To support her, she uses a professional version of Claude as a sort of teaching assistant, fed with last year’s lecture notes, TA solutions and other course materials.
As a reward for showing up to the lecture, Krystal tells students: “Today you’ll get something you won’t find in a book or at home!” She explains: “It’s like a ‘choose your own adventure’ proof. I told students the statement of Menger’s theorem and let them discuss and suggest how to prove it, following their ideas to develop the proof.”
During lectures, students can see Krystal writing live in LaTeX on the screen; and everything is immediately turned into a PDF that becomes the lecture notes. The first half of the lecture, students consider and propose approaches and offer ideas for the proof in plenary. During the break, Krystal feeds their partial solutions into Claude and asks whether they could be completed, and what might need adjusting. Based on that, as a group they try to find a solution after the break. “Without Claude I wouldn’t be able to fix it on the spot because it’s a difficult task and it is not in my specialty area” admits Krystal.
Active design 3: A math lecture with AI and teaching about AI
Krystal also teaches a master‑level lecture in which she starts by asking if students use AI. “They are always a bit scared to admit it,” she observes, “because everyone says ‘no, you can’t!’ Yet hardly anyone shows them how they actually can.”
So Krystal lets Claude generate a full solution to one of her assignment problems. Then the class reads that solution line by line, as a group. “For each sentence, we discussed: ‘Is it true?’ and ‘Does it follow from the previous statement?’ Similar to how you can polish the final draft of a paper.”
Claude’s solution was good, but it also missed a crucial case. And working line by line, the class discovered that. “I heard students say: ‘This is more work than just solving it ourselves’,” says Krystal, “but they really learned to be critical. When Claude makes a mistake you can tell it what it did wrong and then we do that together, on the screen.”
For Krystal, it is important that students learn to be critical of AI output. Ellen agrees and immediately sees potential for earlier years as well: “This could also work for first‑year BSc to get to learn how building a proof works!”
Student experiences: Sharing ideas and enjoying it
One of the biggest worries lecturers have about active lectures is: what if students don’t say anything? Ellen brings this up, as she tends not to have individuals share their answers in front of the group. “I never really have someone share their answer solo, because I think they might be shy. But I do ask every time: does everyone understand?”
Yet Lise and Krystal found that in their lectures, participation grew over time, especially when you build in steps that make speaking less risky. Lise teaches relatively small groups, which helps. But the main factors, she thinks, are pairing up, and consistency. “Coming up with the answer as a pair helped with making it less personal. The first time I did it this way, I would just wait a bit longer and then someone would always answer.”
Lise was positively surprised by how some quiet students transformed once the structure changed. “Some students of whom I thought they weren’t interested really came out of their shell in the team‑up!” Students appreciated her teaching style and described her examples as ‘fun’ and said it ‘felt like playing a game’.
Krystal’s students were also shy at first. But the combination of an exciting challenge and a safe atmosphere changed the dynamic. “My students opened up soon enough. They got so excited about offering their ideas that they forgot to be reserved. They really wanted to share their thoughts.”
Lessons learned: Students appreciate the effort
Krystal was the first to try this AI‑supported, student‑driven proof style, and went in with very few expectations. She is happy with how it worked out. “Several students said that no one ever showed them how to use AI in maths before. The session where the class dissected Claude’s solution line by line got the highest interactivity and engagement I’ve had all term. Students actually came to thank me for my lecture!”
Lise heard from students that her lectures felt as something created especially for them. They had expected standard (boring) lectures and were positively surprised by getting to participate. “They even said I should advertise my lecture style in advance because it would attract more students to actually come to class!”
Ellen noticed something practical: paper. “The first lecture, the first thing I said was: ‘Good morning, take some pen and paper.’ And nobody had it… but the next lecture they came prepared and everyone brought it! For my assignments it is actually much easier to work on paper instead of a laptop, and students enjoy it.” Students also tell her that they appreciate it that at least they really have something to do when coming in at 9.00 on Monday.
Advice for colleagues: Cover less, teach more, have fun and act natural!
All three lecturers advice to cover a bit less, but teach more. Krystal is direct: “Most lectures have too much material, and students are not going to understand all of it anyway. Better use less content that they’ll actually learn”. Thus, focus on core concepts and problem-solving strategies; and check which basics must always be covered to stay in line with other courses. Move details to tutorials.
Also, all three lecturers find their teaching far more enjoyable. Ellen: “If you invest some time to activate your lectures like this, teaching those lectures will be really fun also for yourself.” Lise agrees and states: “Before, in a content heavy lecture, I was sometimes even boring myself…now, not anymore!”
For using complex problems and AI, Krystal further suggests: “Know your material deeply, especially with AI. If you have something less known this may not work for the more difficult solutions.” Stay the expert who can judge and correct AI output.
And don’t worry about participation. “Just pretend that this is normal… and students will follow along”, all three lecturers agree.
Want to try this yourself?
You don’t have to reinvent your entire course. You might start with turning one long explanation into a 10‑minute mini‑lecture plus a short exercise on the same concept. Or, adding one or two structured pair tasks per lecture, with a short plenary discussion. If you want to hear from the lecturers themselves, you can contact Ellen via e.c.m.denouden@uva.nl and Lise via l.stork@uva.nl.
Or try changing just one traditional lecture to a lecture with an example proof like Krystal did. You can find Krystal on her personal website and blog: Krystalguo.com.
You can also find a similar workflow to Lise and Ellen’s in the activity finder: Lectorial.