What Retrieval Practice Actually Looks Like in a Normal Lesson

Most teachers in UK secondary schools have sat through some version of CPD on retrieval practice. You probably remember the slide about Ebbinghaus. You probably remember the line about how testing yourself works better than re-reading. You probably nodded along, because none of it was wrong.

And then you went back to your classroom, and the next time you taught a lesson, you still had to plan it from scratch on Sunday night, and “embedding retrieval practice routinely” did not make the list of things you got round to.

This is not a teacher problem. It is the gap between belief and implementation, and it is doing more damage to retrieval practice than any of the actual debates about it. Most teachers accept the evidence. Most teachers want to use it. Most teachers cannot quite see what it looks like at 9:15am on a Tuesday with thirty Year 9s, a marked book scrutiny on Friday, and the photocopier broken.

What retrieval practice is, in one paragraph

Retrieval practice is recall from memory. That is it. Not re-reading. Not copying. Not highlighting. Pupils trying to remember something without looking it up. The original studies on the testing effect — most usefully Roediger and Karpicke (2006) — showed that the act of trying to retrieve information strengthens memory more than re-studying the same material for the same length of time. We have written separately about the science behind why pupils forget if you want the longer version. For the rest of this post, that one paragraph is enough.

What it does not have to look like

Before the examples, it is worth clearing some misconceptions, because most teachers who say they “are not doing retrieval practice” are doing it badly defined rather than not doing it at all.

Retrieval practice does not require:

• a formal test
• a quiz platform or any technology
• a new printed resource
• a separate lesson
• mark-able output
• preparation time beyond knowing what you taught last week

If a definition of retrieval practice rules any of those things out, the definition is too narrow. It is a category of activity, not a specific format. Almost any task where pupils try to recall something from memory, without their notes, counts. The job is to make that kind of task routine.

Five things retrieval practice actually looks like in a lesson

These are not the only five techniques. They are the five that work in a normal classroom, take little to no preparation, and can be used across phases and subjects without modification.

1. Brain dump

Two minutes. Blank page or back of the exercise book. The instruction is: “Write everything you can remember about [topic from last week / last term / start of this unit].”

That is the whole technique. No prompts, no scaffolding, no template. The blank page is doing the work — it forces pupils to retrieve structure as well as content. You will see who has retained the topic and who has not within ninety seconds of looking around the room.

The version of this that fails is the one with too much support — partial sentence stems, key word lists pre-populated, fill-in-the-blanks. Those turn it into recognition, not recall. Recognition is much easier than recall, which is why pupils prefer it, and which is also why it does much less to strengthen memory.

2. Closed-book questions

Three questions at the start of the lesson, written on the board or said out loud. No notes, no books open. The questions cover something from last lesson, last week, and last term. Five minutes including the answers.

The point of the three different time horizons is spacing. If you only retrieve from yesterday, you reinforce short-term memory. If one of the three questions is always reaching back several weeks, you are doing spaced retrieval without having to plan a separate spaced retrieval programme.

This is the technique that requires the least preparation and produces the most consistent benefit. The bar to use it is that you know what you taught last lesson, last week, and last term. That is it.

3. Verbal cold call

You ask the question, you choose the pupil, the pupil answers, the class confirms or corrects. No hands up. No notes.

Cold calling has acquired a slightly heavier reputation than it deserves — done well, it is just retrieval practice with the additional benefit of distributing thinking across the room. The thing that makes it work as retrieval is the no-notes part. The thing that makes it work pedagogically is choosing the pupil rather than waiting for volunteers, because the pupils with their hands up are usually the ones who least need the practice.

This one is genuinely free. It costs nothing and adds nothing. You probably already do it occasionally. The shift is doing it routinely, with the explicit understanding that you are using it to retrieve, not just to check answers.

4. Retrieval grid

A simple grid of topics from across the unit or term — six to nine boxes — with one question per box. Pupils fill in what they can remember in five minutes. Differentiation happens automatically, because pupils answer what they know.

The grid is the only technique on this list that benefits from being prepared once and reused. You can build a grid for a unit and then re-run it three or four times across the term, with pupils using a different colour pen each time. They can see what they could not retrieve in week two but could in week six. So can you.

The trap with retrieval grids is making them too pretty. They do not need to be designed. A 3×3 table in Word is genuinely fine. The cognitive science does not care what the grid looks like.

5. Exit task

One question at the end of the lesson. Not on what you taught today — on something from earlier in the unit or term. Written, two minutes, hand the slip in or write it on a sticky note.

The reason this is different from a normal plenary is the time horizon. Most plenaries ask “what did we learn today?”, which is a recognition task with the lesson still fresh. An exit retrieval task asks pupils to reach back, which is what produces the strengthening effect. It also gives you a quick read on what has stuck from earlier teaching, which is useful for planning next week.

What makes retrieval practice work — and what undermines it

If you only take one thing from the section above, take this: frequency matters more than intensity.

A two-minute brain dump three times a week does more for memory than a thirty-minute retrieval lesson once a half-term. The whole point of the testing effect is that retrieving is the active ingredient — the stronger the retrieval load, the better, but the more often you reach for the same knowledge, the more durable it becomes.

The most common way schools undermine retrieval practice is treating it as something that happens at the start of term and before assessments. That is not retrieval practice. That is revision. Retrieval works when it is woven into the routine of every lesson, in small, low-stakes amounts.

The second most common way schools undermine it is by changing what is being retrieved. If the knowledge moves — different worksheets, different scheme of work versions, different vocabulary lists between teachers — pupils do not build a stable thing to retrieve. They build a series of half-versions of overlapping things, and retrieval becomes guessing.

This is why the resource question matters more than it looks.

The resource question

Pupils retrieve more reliably when they are clear about what they are expected to know. That is not a teaching problem — that is a curriculum-design problem, but it shows up as a teaching problem in your lesson because retrieval activities work better or worse depending on how stable the underlying knowledge has been.

In practice, this is why a well-designed knowledge organiser makes retrieval easier to run. Not because the organiser is doing the retrieval — pupils should be retrieving from memory, not from the organiser — but because the organiser defines, clearly and consistently, what they are expected to retrieve. The organiser is the source. The retrieval happens from memory. The two roles need to stay separate, which is the most common implementation mistake schools make with organisers and the one most worth knowing about.

If your school is having the broader conversation about how to structure reference material for retrieval, the knowledge organisers overview is the right place to start.

For the lesson on Tuesday, though, the answer is simpler. Pick one of the five techniques above. Use it three times this week. Notice what your pupils can and cannot recall. Adjust nothing else.

That is what retrieval practice looks like in a normal lesson.