You're previewing the Bennett demo family. Numbers, schedule, and kids are sample data.Start your own
Loading page…
Read it again. Then again. It all looks familiar, so you must know it — and that feeling is the single most reliable way to walk into an exam having learned almost nothing. Researchers have been measuring this since 1885.
You read the chapter. Nothing sank in, so you read it again. The second time went faster. The third time you were skimming, because you already knew what the next paragraph was going to say.
You went to bed feeling reasonably prepared.
The next morning it was gone. Not all of it — but the smooth, complete, I've-got-this feeling from the night before had no cash behind it.
Here is the part worth knowing: nothing went wrong. That is what re-reading does. It has been measured, repeatedly, for well over a century, and it comes out the same way every time.
The third time through a page, the words go by easily. Nothing surprises you. Your eyes never snag.
That smoothness has a name — fluency — and it is real. What it measures is how easily text passes in front of you. It is not a measurement of what you could produce with the book shut.
But you have no separate meter for the second thing. So your brain hands you the reading you do have, and fluency feels, from the inside, exactly like knowing.
And notice what a test actually asks. Nobody hands you the page and asks whether it looks familiar. You are asked to produce it out of an empty room. Re-reading practises recognising. The exam wants retrieving. Those are different skills and you have been practising the other one.
Re-reading trains you to recognise a page. Remembering means producing it when the page is not there.
In 2006 two psychologists, Henry Roediger and Jeffrey Karpicke, gave students a short science passage to learn.
One group read it, then read it again, and again, and again.
A second group read it once — and then, with the passage taken away, wrote down everything they could remember. Then did that again. Same total minutes with the material. The only difference was whether those minutes were spent reading it or retrieving it.
Then everybody was tested. Some students five minutes later. Some a week later.
How much each group could recall, as a percentage. Left pair: tested five minutes after finishing. Right pair: tested a week later. Read the left pair first, then the right, and watch what happens between them.
| Study method, and how long before the test | Recalled |
|---|---|
| Reread · 5 min | 83 |
| Retrieve · 5 min | 71 |
| Reread · 1 week | 40 |
| Retrieve · 1 week | 61 |
Look at the two pairs as pairs.
Five minutes out, the re-readers are ahead — 83 against 71. A week out, they have dropped to 40 while the retrieval group is sitting on 61.
Same passage. Same amount of time spent. The method that looked better in the first pair is barely worth two thirds of the other one by the second.
Millions of students re-read, and they are not stupid. Looking at those four bars, what best explains why the habit survives?
When you look at the page, the answer is supplied. You never have to find it.
When you shut the book and try to write down what was in it, you have to find it, and finding it is a completely different action. Sometimes it is horrible. You sit there with four of the seven things and no idea where the other three went.
That horrible bit is the part that works.
Honest caveat, because this channel does not do neuroscience fairy tales: researchers still argue about the precise reason retrieval helps so much. What is not argued about is that it does. It has been found in labs, in real classrooms, with vocabulary, with maps, with medical students, and it holds up.
The moment something is hard to drag out of your head is the moment the practice is happening. Reading the answer skips that moment entirely.
In a follow-up experiment the students were also asked, before any test, to predict how much they would still remember a week later. What happened to those predictions?
Retrieving beats re-reading. There is one more thing, and it is nearly free: when you do those retrievals.
Four goes at a vocabulary list in one evening, or the same four goes across four days. Identical total minutes. The four days win, and it is not close. A review pulling together 254 separate studies and more than 14,000 people found the spread-out version ahead across the board.
And now the pattern from earlier repeats itself exactly. Cramming all four goes together feels great, because by the fourth go in one evening you have the list cold. Come back on Thursday and half of it has leaked away, which feels like going backwards.
It is not going backwards. The forgetting between sessions is what gives you something to retrieve, and retrieving is the thing that works.
Both of the methods that win feel worse while you are doing them. That is not bad luck — the difficulty is where the effect lives.
People have been running this experiment, in one form or another, for about 140 years, and it keeps landing in the same place. Put these in the order they happened.
Tap whichever you think happened next
You learned Tuesday's list properly — by Tuesday night you had it perfectly. On Friday you try to write it out and manage about half. What has Friday actually told you?
None of this needs an app or a special notebook.
One bonus trick that costs nothing: turn every heading into a question before you read the section, then answer it from memory afterwards. You have converted a page into a test, which is the whole idea.
And the thing to stop doing: highlighting. It feels productive, it leaves a satisfyingly colourful page, and when researchers graded ten study methods on the evidence it came out near the bottom alongside re-reading. Both put the answer in front of you. Neither ever makes you find it.
You’re previewing as a parent — nothing here is recorded.
Roediger, H. L., & Karpicke, J. D. (2006). 'Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention.' Psychological Science, 17(3), 249-255. Experiment 1: students studied a prose passage either by repeated study or by repeated testing; final recall was 83% (repeated study) vs 71% (repeated testing) after 5 minutes, and 40% vs 61% after one week. Same paper, Experiment 2: students also gave predictions of their own later recall. The repeated-study group predicted they would remember more, and recalled less at the one-week test — the metacognitive illusion this episode is built around. Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). 'Improving Students' Learning With Effective Learning Techniques.' Psychological Science in the Public Interest, 14(1), 4-58. Ten techniques graded on the evidence: practice testing and distributed practice were rated HIGH utility; rereading and highlighting were rated LOW. Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). 'Distributed practice in verbal recall tasks: A review and quantitative synthesis.' Psychological Bulletin, 132(3), 354-380. Reviewed 254 studies covering more than 14,000 participants; spaced practice outperformed massed practice. Ebbinghaus, H. (1885). 'Über das Gedächtnis' — the first systematic measurement of forgetting over time, using nonsense syllables and himself as the sole participant. Gates, A. I. (1917). 'Recitation as a factor in memorizing.' Archives of Psychology, 6(40). Schoolchildren who devoted the larger share of their study time to reciting from memory rather than re-reading recalled substantially more. Spitzer, H. F. (1939). 'Studies in retention.' Journal of Educational Psychology, 30(9), 641-656. Over three thousand sixth-grade pupils in Iowa; recall was tested at varying delays, and an early test measurably reduced later forgetting. The precise mechanism behind the testing effect is still debated in the research literature; the size and reliability of the effect itself is not. This episode says so rather than offering a tidy neuroscience explanation.