Mind & Psychology · Memory & learning
Why testing yourself beats rereading
Why pulling an idea out of memory builds it better than reading it again, and how to combine recall, feedback and well-timed gaps.

What you will learn
- Tell retrieval practice apart from rereading
- Explain why ease in the moment doesn't prove you'll remember later
- Describe a recall-and-check cycle
- Tell explaining an idea apart from repeating its words
Lesson 1 of 4 · 4 min
Why rereading feels better than it works
You read an explanation twice and it makes complete sense. Then someone asks you to explain it, and half of it won't come. Where did it go?
It didn't go anywhere. While the page was open you were recognizing the idea, which only takes a feeling of familiarity. Explaining it means pulling it out of your own memory, and that's a different job. Doing that job on purpose is called retrieval practice.
Rereading hands you the information again. Retrieval makes you produce it. Both use the same material, and they ask your memory for different work.
In 2006, Henry Roediger and Jeffrey Karpicke tested the two against each other. Students read two short passages. One they then read again. The other they put away and wrote down from memory, with no feedback.
Tested five minutes later, rereading won, 81% to 75%. Tested a week later, recalling won, 56% to 42%. A test after two days went the same way.
So rereading wins the test you take right away and loses the ones that come later. These were college students and short passages, so trust the pattern more than the exact numbers.
A second experiment in the same study exposes a trap. One group studied a passage in four sessions and read it about 14 times in all. Another group read it about 3 times, then took three recall tests. A week later the first group remembered 40% of it and the second group 61%.
Before leaving, both groups had been asked how well they'd remember. The rereaders were the more confident group. Fourteen readings make a text feel thoroughly known, and that feeling says little about what you'll have in a week.
None of this means study should be confusing or needlessly hard. First meet the material and understand it. Then close the page and try a recall you can manage, even a single sentence.
If nothing comes back, look at the explanation again and find what was missing, then try once more. The effort that helps is effort spent reaching for the idea. A blank tells you where to look next. It isn't a verdict on your ability.
Try to recallWhy can rereading feel more successful today yet lead to weaker recall next week, as it did in Roediger and Karpicke's prose experiments?
Material in front of you can feel familiar and boost your confidence without making you retrieve anything. In the prose experiments, rereading helped on the immediate test, but repeated recall did better on the delayed test. How easy something feels now and how well you'll remember it later measure different things.
Rereading feels like learning because recognizing is easy. So once you understand an idea, close the page and say it back before you check. Students who did that kept 61% of a passage for a week. Students who read it 14 times kept 40%.
Lessons in this topic
Why rereading feels better than it works
See why recognizing an idea on the page isn't remembering it, and what happened when students who reread were tested against students who recalled.
What to do after you've tested yourself
Learn why recall needs a check afterward, and how explaining an idea in your own words makes it stick better than repeating it.
How long to wait before you review
Learn why practice spread over time beats practice crammed into one sitting, and how the best gap depends on how long you need to remember.
Why you draw a blank, and what kind of blank it is
Tell a memory that was never stored from one that's stored but blocked, and work out whether old or new learning is getting in the way.
Three quick stories from this topic
The best gap before a review
To remember a fact for a year, the best time to review it is a few weeks later
More than 1,350 people learned facts, reviewed them once and were tested up to a year on. For a test in a week, the best gap before the review was a day or two. For a test in a year, it was 5% to 10% of the year.
Too soon and too late
Lengthening the gap before the review first raised the final score, then slowly lowered it.
Distributed practice
Spreading study sessions over time is called distributed practice. The best spacing depends on how long you need to remember.
The penny nobody can draw
Asked to draw a US penny from memory, most adults placed fewer than half its features
In a 1979 study, 20 American adults tried. The typical drawing had three of eight features in the right place, and the only person to get all eight was a penny collector. You store what you need to tell a coin from other coins, and little more.
Exposure isn't encoding
OpenStax explains the result as encoding failure. To remember something, you usually have to pay attention to its details and actively process them. Most people encode only enough about a penny to tell it apart from other coins, so the finer details never enter long-term memory and can't be retrieved later.
What it means for studying
Seeing something often doesn't mean you've stored it. Memory usually needs effortful encoding, which means paying attention to details and actively processing them. Repeated exposure without that effort leaves little to retrieve.
Fourteen readings lose to three
Reading a passage 3 times and recalling it beat reading it 14 times, a week later
In a 2006 experiment, one group studied a short passage in four sessions. Another studied it once, then wrote down what they remembered three times. After a week the first group recalled 40% and the second 61%. The rereaders had felt more confident.
The early lead
Five minutes after studying, the rereaders were ahead, 83% to 71%. The lead had reversed by the test a week later.
The testing effect
Taking a memory test strengthens later retention as well as measuring it. Psychologists call this the testing effect.
Sources
- Ways to Enhance Memory, OpenStax, Psychology 2e
- Test-enhanced learning: taking memory tests improves long-term retention, Roediger and Karpicke, Psychological Science; Europe PMC abstract
- Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention (full text), Roediger and Karpicke, Psychological Science 17(3), 2006. Copy hosted by East Carolina University
- Spacing effects in learning: a temporal ridgeline of optimal retention, Psychological Science (PubMed record)
- Problems with Memory, OpenStax, Psychology 2e
- Long-term memory for a common object, Nickerson and Adams, Cognitive Psychology 11, 287-307, 1979. Copy of the article hosted at gwern.net





