LumoTutor
← All posts

Why rewatching a lecture doesn't work (and what to do instead)

Rewatching feels like studying because the lecture gets easier to follow. The research says that feeling is misleading. Here's what the experiments show, and a method that works.

LumoTutor · September 25, 2026 · 6 min read

One student in a study of lecture recordings put it better than any researcher could: "Even though it's like a 45 minute lecture it will take me like two hours to get through it." Another: "I'll have to do the whole thing again and pause it like a hundred times." (source)

If that sounds like your week, the problem probably isn't effort. It's that rewatching is one of the least effective ways to study a lecture, and it happens to feel like one of the most effective. Those two facts are connected.

Rewatching makes the lecture easier, not you better

The second time you watch a lecture, it's easier to follow. You recognize the slides, you know where the argument is going, the professor's examples feel obvious. That ease feels like learning. Mostly, it isn't.

Psychologists call this the fluency illusion, and it has been measured directly. In one experiment, students watched the same recorded lecture delivered two ways: once by an instructor who was smooth and confident, once by one who stumbled and read from notes. Students rated the fluent instructor far higher and predicted they'd learned more. On the test, the two groups scored the same, both immediately and a day later (Carpenter et al., 2018).

A larger study found the same gap from the other side. In a randomized experiment in big intro physics classes, students learned more from active sessions where they worked through problems, but felt they learned less, and rated the passive lecture higher (Deslauriers et al., 2019).

Put those together and you get the trap: the study method that feels most productive is often the one that produces the least, and your own sense of "I get this now" is not a reliable gauge.

The "10-minute attention span" is a myth. The real problem is different

You've probably heard that attention collapses after 10 to 15 minutes of lecture. It's one of the most repeated claims in education, and a review of the evidence behind it found little support for it (Wilson & Korn, 2007).

What is well documented is mind-wandering, and it's worse than a neat 10-minute cliff. In one study of students watching a recorded lecture, their minds were elsewhere on 41% of the random checks researchers made while they watched (Szpunar, Khan & Schacter, 2013). That's not a failure of willpower. Watching is passive, and a passive mind drifts.

Rewatching doesn't fix this, because it's the same passive activity a second time.

The experiment that points to the fix

That same 2013 study is the most useful thing to know about studying from recordings, because of how it was designed.

Students watched a statistics lecture split into segments. After each segment, one group answered a few questions about what they'd just seen. A second group got no questions. A third group got the same material again to review, but no questions: effectively, a rewatch.

The group that answered questions:

  • had their minds wander on 19% of checks instead of 41%
  • took roughly three times as many notes
  • scored 89% on the final segment, against 70% for the others

And the rewatch group did no better than the group that did nothing extra. Seeing the material again didn't help. Having to pull it out of memory did.

The study was small and lab-based, so treat the exact numbers as optimistic. But the direction matches one of the most replicated findings in the field.

Why answering questions works: retrieval practice

Testing yourself is not just a way to check what you know. The act of retrieving something from memory strengthens it more than seeing it again does. This is the testing effect, and it holds up across hundreds of studies: one meta-analysis found a clear advantage for testing over rereading (Rowland, 2014), and another, covering 217 studies, found the effect in real classrooms about as strongly as in the lab (Adesope et al., 2017).

Two details from that research matter for how you study:

  1. The benefit grows the longer you need to remember. Retrieval beats rereading more at a week than at a day. Exams are rarely tomorrow.
  2. It works better with feedback. Checking whether you were right, and why, matters.

There's a timing problem hiding in how most students use recordings, too. More than half of recording views in one study happened within 24 hours of an exam (source). That's cramming, and spacing study out over days beats cramming reliably, across hundreds of experiments (Cepeda et al., 2006).

A better way to study a recorded lecture

None of this means recordings are useless. It means using them as something you work with, not something you absorb. Here's a method built from the research above.

1. Watch once, and stop at every natural break. When the professor finishes a point, a derivation or a slide, pause. Don't rewind yet.

2. Before you continue, say or write what that section said, without looking. One or two sentences. If it's a derivation, try to reproduce the next step before the lecture shows it. This is the move that cut mind-wandering in half in the experiment. It will feel slower and harder than just watching. That feeling is the point.

3. When you're lost, write down the exact question before you rewind. Not "I don't get this" but "why does the dxdx disappear between these two lines?" A specific question tells you exactly which 30 seconds to rewatch, instead of replaying ten minutes and hoping it lands this time.

4. Check yourself for the explanation, not just the answer. When you look up whether you were right, look for why. In a large review of feedback research, explained feedback had a meaningful effect on learning, while simple right-or-wrong feedback had close to none (effect sizes of 0.49 against 0.05, Van der Kleij et al., 2015).

5. If you had something wrong, name the wrong idea out loud. "I thought the chain rule only applied to the outside function" does more than rereading the correct rule. Studies that directly refute a specific misconception beat ones that just explain the right idea (Schroeder & Kucera, 2022).

6. Turn your pause-questions into a quiz for later. Answer them again tomorrow, then again in a week, without rewatching first. Rewatch only the parts you get wrong.

7. Don't use recordings as a reason to skip class. The evidence here is mixed, but one study found that when recordings were introduced, attendance dropped, attendance predicted grades, and watching the recordings didn't make up the difference (Edwards & Clinton, 2019).

What a 50-minute lecture looks like done this way

StepTime
Watch, pausing at 6 to 8 breaks to recall each section~65 min
Write down every "exact question" as it comes upincluded above
Rewatch only the spots behind those questions~10 min
Next day: answer your own pause-questions from memory~10 min
One week later: same questions, rewatch what you missed~10 min

That's about an hour and a half across a week, less than the two hours the student in that study spent grinding through a single viewing, and it's built on the parts that actually move the needle: retrieval, spacing and specific feedback.

The short version

Rewatching a lecture makes the lecture feel easier. It doesn't make you remember it. What works is interrupting yourself to recall what you just heard, getting a real explanation when you're wrong, and coming back to it days later.

It's also why we built LumoTutor the way we did: you pause any lecture at the exact moment you're lost and ask about that moment specifically, instead of rewinding and hoping. But the method above works with nothing more than a pause button and a sheet of paper.

Sources