Closed-Book Reconstruction

The strategy, as published

Exactly as it appears on the card

Closed-Book Reconstruction

Students reconstruct key ideas from memory, then open notes and correct in a different colour. Karpicke and Blunt found this outperforms re-reading. The corrections are the data.

Suggested technology document tool; colour annotation · Informed by Karpicke & Blunt

Part of The Testing Effect in 01 Retrieval Practice, where it sits alongside its neighbouring strategies, the infographic and the thinking for the whole area.

Why you might use it

The problem it may help with

Reconstruction sits between a brain dump and a formal test. Pupils rebuild the key ideas of a topic from memory, then open their notes and correct in a different colour. Karpicke and Blunt found this outperforms re-reading.

The second colour is what makes it more than a memory exercise. The corrections are the data: for the pupil, a precise record of what did not come back; for you, a page you can read at a glance.

What it might look like

One version of it, in a classroom

At the end of a unit on rivers, pupils close their books and write out the processes of erosion from memory. Ten minutes later they open their notes and add what they missed in green pen. Flicking through the books afterwards, the teacher can see the same two processes missing in green across half the class.

Try it

A way to begin

Run it once at the end of a current topic: reconstruct from memory, then correct in colour. The colour matters; without it the gaps disappear into the page.

What to notice

Reading what happens

  • Where the second colour clusters. Whole-class gaps are a reteaching list, written by the pupils themselves.
  • Whether corrections are additions or fixes. Missing knowledge and mangled knowledge need different responses.

Things to think about

Before and while you use it

  • Resist marking the reconstruction itself. The imperfect first attempt is the mechanism, and stakes push pupils back towards playing safe.
  • The check against notes is part of the learning, not an admin step; give it proper time.

The wider thinking

Where the evidence sits

Karpicke and Blunt found rebuilding from memory outperformed re-reading for retention, even though the re-reading felt more productive at the time; this is the testing effect applied to whole topics rather than single questions. The card’s attribution stands as published: Karpicke & Blunt.

The fuller picture, sources included, is in further reading and evidence for Retrieval Practice.

Keep exploring

Where next?

Back to this strategy in 01 Retrieval Practice, across to what you're trying to improve, or into all 144 at once.

Want to take this further?

Support is available

Pedagogy First gives you a starting point. If you want support exploring how approaches like this can work across your school, trust or professional learning programme, that work is available.

Work with Mark