Pedagogy

Retrieval practice

Retrieval practice is the act of pulling information out of memory rather than putting it in. It is one of the most reliably effective learning strategies in cognitive science.

By Philip BellLast updated 26 May 2026
Evidence
Strong
Subject
Cross-curricular
Key stage
All
Citations
4
Lethal mutations
3

Retrieval practice is one of those rare findings where the lab evidence, the classroom evidence, and teacher experience all point the same way. Pulling something out of memory is what makes it stick.

What it is

The act of trying to remember information from long-term memory, without the support of notes or the textbook in front of you. The mechanism is sometimes called the testing effect. It works because each attempt at recall strengthens the memory trace and the cues that lead to it.

It is different from rereading or highlighting, which feel productive but produce smaller gains.

Why it works

Memory is built by use, not by exposure. The first attempt to recall something forces the brain to reconstruct it, which strengthens the network of connections around it. Repeated retrievals, spaced out, build durable knowledge.

Karpicke and Roediger’s 2008 study is the classic citation. Dunlosky and colleagues rate retrieval practice as one of the two highest-utility strategies in their review. The EEF cognitive science review lists it among the most strongly evidenced techniques.

How to use it

A five-minute starter quiz on last lesson, last week, and last term. A brain dump on a blank page covering everything pupils know about a topic. Free recall paired with a vocabulary checklist. Cold call questions on prior content. Multiple choice quizzes where wrong answers are common misconceptions.

The Chalk Retrieval Task tool generates spaced retrieval tasks with icons and keywords if you want a quick weekly routine that varies the format.

When not to use it

Retrieval is useless if there is nothing to retrieve. Pupils need to have encoded the material first, through explicit teaching. Do not use retrieval as a substitute for instruction.

Retrieval Task, Visual Keywords, and Worksheet Creator all support low-stakes retrieval routines.

Evidence

Strong evidence

Robust evidence across age groups and subjects, in both lab and classroom settings. Pulling information from memory strengthens it more than re-reading or highlighting. The effect generalises from word lists in lab studies to classroom learning, with smaller but reliable gains.

Caveats

  • Pupils need to have encoded the material first. Retrieval is useless when there is nothing to retrieve.
  • Effects are larger when retrieval is spaced over time rather than massed in one session.
  • Lab effects translate to classroom but typically with smaller effect sizes.

Populations studied

  • Undergraduates
  • Primary and secondary pupils
  • Older adults

Lethal mutations

Common ways this method drifts from the evidence base. Drawn from Brown and Campione's research on implementation degradation.

  1. 1

    Treating retrieval as a high-stakes test

    Why it matters: Retrieval works best when low-stakes. Anxiety from grading interferes with the very mechanism (effortful recall) that strengthens memory.

    Do this instead: Keep retrieval ungraded. Frame as a chance for pupils to find out what they've forgotten, with no marks attached.
  2. 2

    Showing the answers immediately as a slide while pupils still copy them

    Why it matters: This converts retrieval back into restudy, removing the desirable difficulty that drives the effect.

    Do this instead: Let pupils struggle with the recall first, then reveal answers. The struggle is the mechanism, not a side effect.
  3. 3

    Using only the same question format week after week

    Why it matters: Pupils learn to retrieve in that one format. Recall fails when the exam asks the same content in a slightly different way.

    Do this instead: Vary the format. Free recall, multiple choice, short answer, brain dump, sketch-and-label. Different cues build flexibility.

Try this in Chalk

Related concepts

Questions teachers ask

Is retrieval practice the same as testing?
Sort of. The mechanism is the same, but retrieval practice in lessons is usually low-stakes. A quick five-question quiz at the start of a lesson, a brain dump on a blank page, or a flash card round all count.
How often should I do retrieval practice?
Little and often beats long and rare. A five-minute starter every lesson outperforms a long quiz once a fortnight. Spacing the retrievals out across weeks is even better.
Will pupils find it stressful?
Only if you make it stressful. Keep it low-stakes, unmarked, and frame it as a chance to find out what they have forgotten. The Dunlosky review notes that retrieval practice works across confidence levels.
Try it in Chalk

Retrieval Task

Low-stakes recall task with icons and keywords for spaced practice across a unit.

Open Retrieval Task
Published 26 May 2026. Last reviewed 26 May 2026. Chalk content is reviewed against the evidence at least once a year.