Before you read any further, try something. Think of the last course video you watched, at work or at home. Now name three things it taught you. Not the title, not the topic: three actual points, in your own words. If that came slowly, or not at all, you are in very good company, and the video probably felt perfectly clear while you were watching it. That gap between how clear it felt and what you can produce now is the whole subject of this post. It is written for people learning from video courses alongside a full-time job, who follow every lecture and still find, a week later, that very little of it is usable.
The gap has been measured. In 2019, physicists at Harvard ran an experiment inside a 15-week introductory physics course. In week 12, half the class was randomly assigned to a highly polished lecture and the other half to an active session where students worked through problems themselves, with help on hand. The content was identical. Then the groups swapped. After each class, students rated how much they felt they had learned and took a 12-question test. They felt they learned more from the polished lectures. They scored higher after the active sessions. The lead author, Louis Deslauriers, explained why:
Deep learning is hard work. The effort involved in active learning can be misinterpreted as a sign of poor learning. (Harvard Gazette, 2019)
His summary of the result was blunter: actual learning and the feeling of learning were "strongly anticorrelated". They pointed in opposite directions. The smooth experience felt like progress, and the effortful one produced it.
The reason is not mysterious. When you watch a clear explanation or reread your notes, everything looks familiar, and familiarity feels exactly like knowing. But recognising something while it is in front of you is a different skill from producing it when it is not. Work only ever asks for the second one. Nobody in a meeting hands you the slide before asking why the schedule slipped.
The two-minute check
The fix is not to study more. It is to change the last two minutes of every session, from looking at the material to pulling it back out of memory. Researchers call this retrieval practice, and it is one of the best supported findings in learning research: a major 2013 review rated it one of only two techniques with high utility, and a 2021 meta-analysis of more than 1,600 effects put its effect well above rereading or highlighting.
Did the test at the top of this page come easily? If it did, this is probably already how you study, and the sections below will mostly confirm it. If it did not, the check above is the single change most likely to help.
"Testing myself is for exams. I am learning this for work."
I am not sitting an exam. I need to use this on the job, not recite it. Quizzing myself sounds like school.
It is a fair objection, and the research answers it better than you might expect. Retrieval does not only help you remember for a test. A 2017 meta-analysis of retrieval practice found it improved transfer, applying what you learned to new questions, as well as straight recall. And work is a test, just one without a date. The question from your manager, the decision in a planning meeting, the incident at two in the morning: each one asks you to produce something from memory, without the course in front of you. The two-minute check is rehearsal for exactly that moment.
Picture a project manager who has just watched a lecture on risk registers. Rewatching it would feel productive. Writing three points might produce "owners per risk", "review every two weeks" and a blank where the third should be. The blank is the valuable part: it is the thing they would have stumbled on when the sponsor asked.
"I do not have time, and it feels worse."
I squeeze lectures in between meetings. Rewatching the bit I half followed is quicker, and I come away feeling I have got it. Writing things down from memory just makes me feel slow.
That feeling is the thing the Harvard study measured, and it was wrong. The ease of rewatching is precisely why it feels effective, and precisely why it works less well. The 2013 review rated rereading and highlighting as low utility, not because they do nothing, but because the same minutes spent on retrieval do much more, and because they leave you more confident than you should be.
The time argument also runs the other way. Rewatching a fifteen-minute lecture costs fifteen minutes. The check costs two, and it tells you which ninety seconds of the lecture you actually need to see again. Does that match how your study time goes now? If you rarely rewatch anything, this objection is not yours, and the next one probably is.
"My course has no quizzes, and nobody checks."
The course I am on is just videos. And at work nobody asks whether I learned anything. There is nothing to test myself with.
You do not need anyone else's questions. The section titles in the course outline are ready-made questions: turn "Stakeholder mapping" into "How would I map the stakeholders on my current project?" and answer it on paper. Downloads and slides make a perfect answer key. And explaining an idea out loud to a colleague, or to an empty room, is retrieval with a built-in check, because you hear the moment your explanation runs out.
Where a course does have quizzes, treat them as part of the learning rather than a verdict on you. In one well-known study of video lectures, students who were quizzed between segments let their minds wander about half as often and did better on the final test, although later studies have been more mixed.
What the feeling is for
The feeling of learning is not useless. It is just answering a different question. It tells you whether a lecture was pleasant and clear, which is worth knowing. It does not tell you whether you can use what was in it, and on that question it is often backwards. So when a session felt easy, be a little suspicious of it, and when the two-minute check feels slightly hard, take that as a sign it is doing its job. The test at the top of this page took you less than a minute. Run it on yourself after every lecture and you will know, rather than feel, what you have learned.
Sources: Deslauriers et al. 2019, via the Harvard Gazette | Dunlosky et al. 2013, Psychological Science in the Public Interest | Donoghue and Hattie 2021, Frontiers in Education | Adesope et al. 2017, Review of Educational Research | Szpunar, Khan and Schacter 2013, PNAS | Redifer et al. 2025, on retrieval and complex texts
