Recording Lectures Instead of Live Teaching
The basic idea is straightforward enough, though the execution usually goes wrong. Instead of standing in front of a classroom explaining how to find the derivative of x squared, you make a video about it. Students watch that video before coming to class. The class time then becomes about doing problems together while you walk around and catch the mistakes people are making. That's the model. Everything else is details. I've been doing this for about eight years now. I started because I realized my lectures were mostly me talking at people who were either already asleep or frantically copying down equations without understanding anything. The first version of the flipped calculus class I ran had roughly forty percent completion on the pre-class videos. That was my fault, not the students' fault. I made videos that were too long and too dry. After I cut them down to eight to twelve minutes and stopped reading off slides, completion jumped to about eighty percent. The difference came from actually showing the work on a document camera rather than narrating over PowerPoint bullet points.
The Flipped Math Calculus Revolution How Video Lectures Are Transforming Student Mastery Of Derivatives And Integrals
One thing most people don't tell you about teaching derivatives through video is that silence matters. When you're solving a problem like finding the derivative of a composite function using the chain rule, pausing for three seconds after setting up the problem lets students write their own attempt before you reveal the answer. I learned this the hard way. My first batch of videos were twenty minutes of non-stop talking. Students would pause, rewind, and get confused about where the explanation was going. Once I started embedding deliberate pauses into the videos, engagement went up and the number of students who said "I understood the video but couldn't do the homework" dropped significantly. The chain rule is where this model shows its real value. Students can watch the video at half speed, pause to work through each step, and rewatch the part where you set up f of g(x) before taking the derivative. In a live classroom, you explain it once and move on. If someone misses it, they've missed it. With video, they control the pace. The tradeoff is that some students treat the video like background noise and only engage when there's a quiz attached to it. I grade the pre-class videos now with short checks that require a screenshot of the student working a problem on paper, not just clicking "I watched this." That eliminates the passive viewing problem for most of the class. For integrals, the flipped approach hits a different wall. Integration is fundamentally harder to teach through a lecture video because there's rarely a single path to the answer. You can show one method for integrating by substitution, but students still have to learn when to apply it versus integration by parts versus partial fractions. I handle this by making a series of shorter videos, each focused on recognizing the structure of an integral, followed by a class session where the only activity is comparing different integrals and deciding which method applies. The meta-cognitive work of choosing the right technique is something you can't really put in a video. That has to happen in person.
Here's an edge case that tripped me up for a full semester. I was teaching implicit differentiation and had a video where I solved for dy/dx step by step. During class, I asked students to work on a problem that required implicit differentiation combined with the quotient rule. Half the class froze. Not because they didn't know implicit differentiation or the quotient rule separately, but because the video never showed them what happens when you combine both techniques in a single problem. The gap was invisible to me as the instructor. I had built up the concepts separately but never demonstrated their intersection. The workaround was simple: I added one "challenge variant" video per topic that showed the combined technique, and I put it in the module after the foundational videos but before the class meeting. That single addition closed the gap. I still occasionally design videos that are too modular, which leaves students unprepared for the problems that actually appear on exams. The biggest practical limitation of this model is something nobody talks about much: it assumes students have reliable internet and a quiet place to work during the week. I've had students in every semester who watched the videos on their phone during their commute because they worked a job and couldn't be home by seven in the evening. They heard the audio but couldn't see the handwritten work clearly. For those students, I record a second version of each video in a slightly simpler format with larger handwriting and slower pacing. It's extra work on my end, but the alternative is losing those students entirely during the unit on related rates, which is already the hardest topic in AP Calculus. Another honest limitation is that the flipped model doesn't help students who haven't developed basic algebra skills. I had a semester where roughly a third of the class struggled with the pre-class videos because they couldn't factor polynomials or simplify rational expressions. No amount of recalibrating video length or adding pauses would fix that. Those students needed remedial algebra support alongside the calculus course, and the flipped format doesn't create that support by itself. I recommend pairing video-based instruction with a weekly algebra review session for classes that have a significant scatter in foundational math skills.
Get the Full Details
What the Videos Actually Need to Contain
A useful calculus video for the flipped classroom is not a recording of a lecture. It's a demonstration. You show the problem. You think out loud while solving it. You make mistakes deliberately and correct them. Students learn more from watching you discover a sign error than from watching you get it right the first time every single time. The videos I get the best results from are the ones where I struggle with a limit evaluation, realize I've been assuming continuity without checking it, and then go back to verify each condition. That process takes about ten minutes and teaches more than any perfectly polished solution would. I organize my content by topic in modules. Each module has two to four videos, each under fifteen minutes, followed by a problem set that takes about twenty minutes to complete during class time. The problem sets are the same ones I would have assigned as homework, except now I'm available to help while students work through them. The net effect is that homework completion rates went from roughly sixty percent in the traditional setup to about ninety-five percent in the flipped setup. The quality of the finished work improved as well, which suggests that students aren't just copying answers anymore. If you're planning to build this yourself, the tooling is trivial. A document camera or a tablet with good handwriting recognition software works fine. I use a Wacom tablet with one-note and export the recordings as MP4 files. Students don't need special software to view them. Upload everything to your learning management system and add the short video checks to track whether anyone is actually watching. The actual technical barrier to entry is near zero. The difficulty is in making videos that students will watch, which comes down to editing discipline, not technology.
Where This Model Breaks Down
Not every calculus class benefits equally from flipping. Advanced placement courses with students who already have strong motivation tend to do well. Remedial college courses with students who struggle with time management often don't, unless the institution pairs the flipped format with mandatory study sessions or supervised video viewing periods. I've seen schools try to flip entire gateway math sequences without providing any structure for student accountability, and the results were consistently worse than traditional lectures. The videos got ignored, class time became unstructured, and learning outcomes declined. The approach also doesn't work well for proof-based courses or classes that rely heavily on synchronous discussion. If your calculus course includes rigorous epsilon-delta proofs, the flipped model requires very different video design than it does for computational topics. I don't recommend flipping that content unless you're prepared to invest significant time in scripting and rehearsing the videos, because the cognitive load of following a proof in real time is much higher than following a computational demonstration. The core insight is that flipping moves the transmission of information outside of class and reserves class time for the part of learning that actually requires a teacher present. That's a significant shift and it doesn't automatically mean better outcomes. It means the outcomes depend on whether students will engage with the videos and whether you have enough class time to turn that engagement into actual skill. When both conditions are met, derivatives and integrals become significantly more accessible than they are in a standard lecture format. When they're not, you've just added another layer of work for everyone without changing the result.
