Retrieval practice has become one of the most widely adopted teaching strategies in UK schools. The EEF's Cognitive Science Approaches in the Classroom review found that more than 85% of teachers describe cognitive science strategies as central to their practice, and retrieval is the most-used of those strategies. Walk into most secondary classrooms in the country and you'll see a low-stakes quiz, a do-now recall task, or a "five from last week" question on the board.
But the EEF's review also flagged something important. Retrieval practice works, but only when it's done well. Done badly, it becomes a routine that takes up lesson time without producing the memory gains it's supposed to.
In my own classes, I routinely begin every lesson with a short low-stakes retrieval quiz. It's usually 5 questions, and I intentionally mix the difficulty. I'll start with 1 or 2 accessible questions to build confidence, then move into more challenging recall that touches on the foundational knowledge needed to access today's topic. It's become a routine, a class settler, and a way to shape the lesson ahead. There have been times where student performance on this task makes the rest of my plan go out of the window. I'd assumed they'd know something, they didn't, and now I need to reteach it before moving on.
This article sets out what retrieval practice actually is, why it works, and how to do it in a way that genuinely improves learning rather than just ticking the cognitive science box.
What Retrieval Practice Is (and Isn't)
The EEF defines retrieval practice as recalling information from memory with little or minimal prompting. The key word is recall. Re-reading notes, highlighting a textbook, or watching a video again is not retrieval practice. It's review, and the evidence is clear that review alone produces much weaker memory gains than active recall.
The foundational research comes from Roediger and Karpicke (2006), who showed that students who studied a passage and then took a recall test remembered significantly more a week later than students who simply re-read the passage the same number of times. The effect is now known as the testing effect, and it has been replicated across age groups, subjects, and continents.
What makes the testing effect powerful is the act of struggling to remember. The cognitive effort of pulling information out of long-term memory strengthens the memory trace far more than passively reviewing the same content. That difficulty is the point, not a problem to design around.
Why Retrieval Practice Works
Three things happen when a pupil retrieves information from memory.
- The memory trace is strengthened. Each successful retrieval makes future retrieval easier and more durable.
- Gaps in knowledge become visible. Pupils notice what they can't recall, and so do you, which makes retrieval a powerful diagnostic tool as well as a learning tool.
- Knowledge becomes more flexible. Repeated retrieval in different contexts helps pupils transfer what they know to new problems, rather than just regurgitating it in the form they first met it.
This last point matters because there's a common criticism that retrieval practice is just rote learning. The evidence suggests otherwise. When retrieval is varied (different question formats, different contexts, spaced over time), pupils build the kind of flexible, durable knowledge that supports higher-order thinking later.
The Four Conditions for Retrieval Practice to Actually Work
Drawing on the EEF review, four conditions determine whether your retrieval routine is genuinely improving learning or just filling lesson time.
1. The difficulty has to be right
If retrieval is too easy, pupils don't have to think hard, and the memory benefits disappear. If it's too hard, they get nothing right, give up, and learn nothing. The sweet spot is what cognitive scientists call desirable difficulty. These are questions pupils can answer with effort, but not without it.
A common mistake is asking pupils to recall material they've only just covered. Recall is much more powerful after a delay (a few days, a week, or longer) because the act of pulling it back from semi-forgotten storage is what builds durable memory.
2. Pupils need feedback on what they got wrong
Retrieval without feedback teaches pupils nothing about gaps in their knowledge. The feedback doesn't have to be elaborate. Going through the answers as a class, or having pupils self-mark with the correct answers visible, is enough. What matters is that misconceptions are corrected before they harden.
3. It has to be low-stakes
The whole point is that pupils try to recall, get some things wrong, and learn from the errors. If retrieval is graded or contributes to formal assessment, pupils start avoiding the discomfort of being wrong, which is exactly the discomfort that produces learning. Low-stakes routines create the conditions for productive struggle.
4. Pupils need to know why they're doing it
A widely-cited insight from cognitive science research is that pupils benefit from understanding why a teaching strategy is being used. Telling a class explicitly that low-stakes quizzes are designed to help things stick in long-term memory, rather than to test or punish them, increases buy-in significantly.
You'll always get the question, 'What's the point in this? We did some of these last week.' I explain to my students that repetition and practice is the key, just like in their favourite sport. You don't stop shooting practice in basketball because you scored once last week. The same principle applies to learning, you keep retrieving until it sticks.
Practical Retrieval Practice Formats
The format matters less than the four conditions above, but variety helps. Here are the formats that work well across subjects.
Low-stakes quizzes
Five to ten short questions at the start of a lesson, drawn from the previous lesson, the previous week, and earlier in the term. The mix is important. It builds in spacing automatically.
Brain dumps
Pupils get a blank page and write down everything they remember about a topic in a fixed time (usually 3 to 5 minutes). Hugely effective for revealing what's actually in memory rather than what pupils think is in memory.
Cold call questioning
Asking specific pupils to recall specific things, rather than taking volunteers. The cognitive load of 'it could be me' keeps the whole class engaged in retrieval, not just the pupil being asked.
Flashcards
Self-testing with flashcards is one of the most evidence-supported retrieval methods, particularly for vocabulary, definitions, and discrete facts. The key is that pupils have to attempt the answer before flipping the card. Re-reading a stack of flashcards isn't retrieval practice.
Spoken retrieval
Less common but increasingly recognised. Pupils explain a concept aloud, summarise a topic verbally, or answer questions in spoken form. This adds an oracy dimension to retrieval, which has its own benefits, and is the basis of how QuizCards is designed. Pupils answer flashcard questions out loud, with their answers assessed against the expected response. For teachers looking for spoken retrieval routines, this is one option among several.
Common Mistakes to Avoid
A few traps worth flagging, drawn from the EEF guidance and from teacher experience.
- Mandating retrieval at the start of every lesson without thinking about content or difficulty. The routine becomes empty.
- Using retrieval only for low-level facts. It works for higher-order content too, but the questions have to be designed well.
- Making it high-stakes. Pupils stop taking productive risks.
- Skipping the feedback step. Without correction, errors get rehearsed alongside correct answers.
- Not spacing it out. Asking pupils to recall content from earlier today is much weaker than asking them to recall content from last week.
Retrieval Practice and Other Cognitive Science Strategies
Retrieval rarely works alone. The EEF review groups it with two related strategies that compound its effect.
- Spaced practice. Reviewing the same content with gaps in between produces stronger memory than reviewing it all at once. A retrieval quiz that mixes content from this week, last week, and last month is doing both retrieval and spacing.
- Interleaving. Mixing different topics or types of problem in one practice session, rather than blocking them, supports flexible recall. Useful for maths, languages, and any subject where pupils need to discriminate between similar concepts.
A retrieval routine that incorporates spacing and some interleaving is doing far more work than a single-topic recall starter.
A Starting Point for the Coming Term
If you're new to retrieval practice or looking to sharpen what you already do, three things are worth committing to.
- A weekly low-stakes quiz that mixes content from the past week, the past month, and the past term.
- One brain dump per topic, used as a diagnostic before you teach a key idea.
- A short, explicit conversation with each class about why you're doing retrieval and how it helps them learn.
You don't need a new platform, a new resource, or a new template. You need consistency, the right level of difficulty, and feedback. Everything else is detail.
FAQs
What is retrieval practice?
Retrieval practice is the strategy of recalling information from memory with little or minimal prompting. It strengthens long-term memory more effectively than re-reading, reviewing, or re-watching the same material.
Are quizzes the same as retrieval practice?
Quizzes are one form of retrieval practice, but not the only one. Brain dumps, cold call questioning, flashcards, and spoken recall are all retrieval. The format matters less than whether pupils are genuinely recalling from memory.
How often should I use retrieval practice?
The evidence supports regular, spaced retrieval rather than daily intensive use. A weekly quiz mixing content from different points in the term tends to be more effective than a daily recall task on yesterday's lesson.
Should retrieval practice be graded?
No. The evidence supports low-stakes retrieval, where pupils can be wrong without consequence. Grading retrieval undermines the productive struggle that makes it work.
Does retrieval practice work for higher-order thinking, or just facts?
It works for both, but the questions have to be designed well. Recall of facts is the easier application. Recall of concepts, frameworks, and processes requires more careful question design but produces the same kind of memory benefit.
It would be great to hear how retrieval practice is working in your classroom. Share your routines or what's not landing yet in the comments below or get in touch.
Comments