Why children remember what they work out and forget what they’re told
Being told is the fastest way to transmit an answer and the slowest way to install one. There is a large, boring, extremely reliable body of evidence behind that — and one important limit on it.
The retrieval effect, and why nobody believes it
In a well-known experiment, students who studied a passage and then tested themselves remembered dramatically more a week later than students who studied it repeatedly. The interesting part: the re-readers were confident they had learned more. They felt fluent and were wrong.
This is the single most practically useful finding in learning research, and it maps directly onto a child at a puzzle. Being shown the solution feels like understanding. Producing it — badly, slowly, with two false starts — is what stays.
The second free upgrade: come back to it
Learning that is spread out beats learning that is bunched, for the same total time. Come back to the same idea a week later, a little harder, and it moves from fragile to durable. That is why our curriculum spirals rather than marches.
The failed first attempt is not wasted time. It is what makes the explanation mean something.
The limit: pure discovery does not work
It would be convenient to conclude that children should be left to figure everything out. The evidence says otherwise, clearly. Unguided discovery learning consistently underperforms guided instruction — children flounder, form wrong models and take much longer to get nowhere.
What works is the middle: let them attempt it first, let the attempt fail informatively, then guide — and then make them try again. The attempt before the explanation is what makes the explanation land.
Pure discovery. Consistently beaten by guided approaches in controlled comparisons.
Explanation before attempt. Comfortable, fast, and mostly forgotten.
The failed first attempt is not wasted time. It is what makes the guidance mean something.
Questions parents actually search for
Why does my child forget what they were taught last week?
Because being told is a weak way to install anything. Long-term memory responds to producing and retrieving an answer, and to meeting the same idea again after a gap — neither of which happens when something is covered once and moved past.
Is discovery learning better than being taught?
No, not in its pure form. Unguided discovery reliably underperforms guided instruction. The effective pattern is an attempt first, then guidance, then another attempt — which keeps the benefit of struggling without the cost of floundering.
What is the testing effect?
The finding that retrieving information from memory — self-testing, explaining, producing an answer — improves long-term retention far more than re-reading, even though learners usually believe re-reading works better.
How should we practise at home?
Little and spread out. Five minutes of “can you still do this?” a week later beats an hour of drilling on Sunday, and asking them to explain it to you is one of the strongest moves available.
One finding a week, in plain words, with its caveat attached
That is the whole newsletter. Two minutes on a Friday, nothing sold, unsubscribe in one click.