It’s 11 PM, the exam is tomorrow, and you’re on your fourth pass through Chapter 7. Half the page is highlighted. The words look familiar. You feel ready. Then the paper lands on the desk and the chapter has gone.
That feeling of being ready, and then not being ready, has been measured. In 2006 two psychologists at Washington University in St. Louis, Henry Roediger and Jeffrey Karpicke, set out to test it. What they found is one of the most cited results in the study of learning, and it is simple enough to use tonight.
The study, in plain terms
Roediger and Karpicke gave university students short prose passages to learn, the kind of factual reading you meet in a textbook. Some students spent their study time reading the passage again. Others read it once and then spent the same time writing down everything they could remember, from a blank page, without looking back at the text. Nobody told the recall group what they had got right or wrong.
Then everyone took the same final test: write down all you can remember of the passage. Some took it five minutes later, some two days later, some a week later.
What they found
In the first experiment, rereading won the short race. Five minutes after studying, the students who had reread the passage recalled 81% of it; those who had recalled it instead managed 75%. Two days later the order had flipped: 68% for the recall group against 54% for the rereaders. After a week it was 56% against 42%.
The second experiment pushed further. Three groups each had four short study periods. One group read the passage in all four. One read it three times and recalled it once. One read it once and recalled it three times.
| How they studied | After 5 minutes | After a week | How well they expected to remember it (1–7) |
|---|---|---|---|
| Read it four times | 83% | 40% | 4.8 |
| Read it three times, recalled it once | 78% | 56% | 4.2 |
| Read it once, recalled it three times | 71% | 61% | 4.0 |
Look down the last two columns together. Over the week, the students who only reread forgot about half of what they had known five minutes after studying: from 83% to 40%. The students who read once and recalled three times lost far less: from 71% to 61%. And they did it while spending much less time looking at the text.
It isn’t just extra study time
A natural objection is that recalling is simply another way of going over the material. The design rules that out. The recall groups spent their time with the text closed, so they were exposed to the passage less, not more, and received no answers to check against. What they did differently was pull the material back out of their own memory. That effortful act of retrieval, rather than seeing the words again, is what made the difference. Psychologists call it the testing effect, or retrieval practice.
One way to picture it: rereading practises recognising the material, which is what the page lets you do. Recalling practises finding it without the page, which is what an exam, a seminar or a conversation asks of you.
The illusion of competence
The most uncomfortable finding is the last column of the table. Before the final test, the students were asked how well they expected to remember the passage in a week. The rereaders were the most confident. The students who had recalled it were the least confident, and they remembered the most.
Rereading makes a passage feel easy. Each pass is smoother than the last, and that smoothness feels like knowing. It isn’t: it is familiarity, and familiarity fades fast. Recalling feels worse because you meet your own gaps, but meeting them is the point. You find out what you actually know while there is still time to do something about it.
Is this one study, or a pattern?
It is a pattern. In 2013, a review in Psychological Science in the Public Interest by John Dunlosky and colleagues weighed ten popular study techniques against the research. Practice testing was one of only two they rated as highly useful; rereading and highlighting were rated low. A 2014 meta-analysis by Christopher Rowland in Psychological Bulletin, pooling 159 comparisons, found that testing yourself tends to beat restudying, with the advantage largest when the final test comes after a delay.
It has limits worth knowing. Most of this research uses students learning set passages, and the benefit depends on real effort: glancing at a question and then at the answer is not recall. Checking what you missed afterwards, against the source, adds to the effect rather than undoing it.
How to use it tonight
- Read once, properly. Understand it the first time instead of planning to reread it.
- Close it and write down everything you remember. A blank page, no notes, no peeking. Saying it out loud works too.
- Then check against the source. Mark what came back, what came back only partly, and what you missed, and look at the missed parts again.
- Leave a gap and do it again. Recall the next day, and again a few days later. Each time, it takes less.
- Trust the effort, not the feeling. If it feels hard, it is working. If it feels easy, test yourself before you believe it.
How Recitare does this with you
Recitare has a practice built on exactly this, called Tell it back. You choose what to test yourself on: a whole course, one topic within it, or a single reading you have added. You write or say everything you remember, from a blank page, and nothing is shown to you until you finish.
Then Recitare checks what you said against the readings you added. You see which ideas came back, which came back only in part, and for each one you missed, the exact passage in your reading and its page, so you know what to go back to. There is no score: the point is to find the gaps, not to grade them.
If you prefer smaller steps, Study Mode asks a question every few paragraphs while you read, and study guides turn a reading into flashcards you answer before you flip.
Try it with your next reading
Add a reading, then tell it back. Free to start, no card needed.
Start freeSources: Roediger, H. L., and Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255. Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., and Willingham, D. T. (2013). Improving students’ learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4–58. Rowland, C. A. (2014). The effect of testing versus restudy on retention: A meta-analytic review of the testing effect. Psychological Bulletin, 140(6), 1432–1463.