Why don’t students remember what they’ve learned?

Curricula and assessment aren’t designed with memory in mind



We’ve all had the experience of cramming for an exam and forgetting most of what we learned within a few weeks or days. In the immediate term, this is actually quite useful, because national exams are often bunched together, sometimes several in the same day. But in the long-run it’s unhelpful: if you asked me to take a GCSE physics exam today, I don’t expect I’d do very well: I’ve forgotten almost all of what I learned. Given the time invested, it seems a waste.


‘In one ear, out the other’

If you’re in teaching, you’ll have had the frustrating experience of seeing a class grasp and understand a concept perfectly in lessons, only to have completely and utterly forgotten it when you mention it later on in the year: they assure you they’ve never heard of it before in their lives.

If you speak languages, you know what it’s like to get rusty; but also that when you reimmerse yourself in the country for some time, it comes back to you far quicker than learning an entirely new language.

Memory works in mysterious ways. I was reminded of this when I was asked the other day by a mystified English teacher: ‘Don’t you find it frustrating when those on English A-level still don’t remember to use full stops properly?’

The thing I most like about cognitive science is that it helps you understand the puzzling things you encounter in education. One mystery is why students seem to forget so much of what they’ve learned. This puzzle has been on my mind for some time. A while back, I wrote of the troubling lack of knowledge I see in schools in disadvantaged areas, and those colleagues of mine in similar schools:

A colleague of mine teaching English in a disadvantaged school found that pupils were under the illusion that the English language was invented in the 1960s and that Shakespeare wrote the Bible. One teacher found her pupils confused over whether Iran and Iraq were the same country; whether Sydney was in California; and whether Henry VIII is the Queen’s son. Another teacher mentions here that 16 year olds couldn’t place their city on a map of Britain, list the four countries that make up the UK, tell the difference between England and Great Britain, or name the date of one significant historical event. Still another teacher and education blogger I know told me that her pupils thought Manchester was in Scotland, Wales was an island and the Romans came from Portugal. When asked the capital of Wales, pupils’ best guess was Scotland. Many couldn’t spot the UK, the US, or China on a map, even ‘the top set Year 10 superstars.’ Political knowledge also seems impoverished: in another teacher’s school, many pupils couldn’t name the Prime Minister. Some had a hazy idea it was Obama. Some said ‘Gordon Blair’. No one could name all three main political parties, or even any other than Labour. Maths seems to suffer from similar knowledge deficit in some schools in disadvantaged communities. A Maths teacher I know in the West Midlands told me his pupils thought you might measure the distance between Liverpool and London in centimetres; one pupil in a Year 10 top set asked him what ‘square it’ meant; a Year 11 student from another set, one day before a GCSE exam, asked what a percentage was. Even History undergraduates know little: when surveyed by one University professor, around 90% of them could not name one British 19th century Prime Minister. If this is what most History graduates do not know, what can we be sure all school leavers, for whom history is not even compulsory, do know?


One comment replied: ‘Don’t forget that just because pupils don’t know something (for example details about Shakespeare), it doesn’t mean they haven’t been TAUGHT it.’


Kris Boulton, Deputy Head of Maths at King Soloman Academy, is turning out to be a bit of a master of memory, especially in the field of mathematics, and he asks:Why do we keep forgetting so much of what we learn? The other day, a Year 10 girl said: “What’s the point of learning this?” What she meant was, in a week we’d be on to a new topic, and she’d forget everything she could now do, so where was the point in learning it? I haven’t yet seen any institutional focus on the importance of building memories. I would like to suggest that we start to think of building long-term memory retention and recall as a separate concern; that we start to put thought and effort into thinking about how we are going to help students remember what they learn from us, that we ask ourselves at the start of planning a lesson, or a unit ‘How am I going to help ensure my students still remember this six months from now, a year from now, two years from now…?’

I’d like to take up Kris’ challenge and use the cognitive science of how memory works to explain why our pupils forget what they’ve learned, and ask in future posts what we can do about it.

How memory works

Cognitive scientists Dan Willingham and Robert Bjork have been thinking about the issue of how memory works for several decades. Their research helps explain how we commit things to memory, and how we can avoid forgetting them.


Willingham asks: ‘What makes something stick in memory, and what is likely to slip away? How can the memory system know what it’ll need to remember later?’ He answers:

‘Given that you cannot store everything away, your memory system lays its bets: if you think about something carefully [and repeatedly) you’ll probably have to think about it again, so it should be stored. If you don’t think about something very much, then you probably wont want to think about it again, so it need not be stored. Your memory is a product of what you think most carefully about. What students think about most carefully is what they will remember.’

Willingham’s model of the mind simplifies cognitive architecture into working memory and long-term memory:


So, why might students forget things they’ve been taught? Willingham suggests a number of reasons:

  1. Attention: you can’t remember things you haven’t paid sustained attention to in working memory.
  2. Storage: you have paid attention, but it hasn’t made it into long-term memory – it never stuck.
  3. Usage: you can’t remember things that no longer reside in long-term memory – they have faded through disuse.
  4. Transfer: your process by which things are drawn from long-term memory is prone to failure: transfer is difficult, because it’s difficult to apply abstractions to new situations.

In short, we don’t remember things because of insufficient focus, time or attention spent on them, and because of insufficient practice, usage, revisiting, consolidation or application.

So, when we grumble as teachers that students don’t use punctuation properly, even though they’ve learned it, we need to ask ourselves: have they really learned it? Have we really taught it with sufficient time, focus and attention? Have we sufficiently revisited it? Have we consolidated it in their minds? Have they mastered it? Have they automated it in their long-term memories?


It’s no good grumbling about pupils’ written ineptitude. Punctuation is complex: even the basics of using full stops requires lots of little chunks of knowledge to be automatic: to decide whether a sentence has a subject and a verb to avoid fragments, and whether a sentence runs on into multiple, confusing independent and subordinate clauses, you need to know all those concepts: what subjects, verbs, independent and subordinate clauses are! It may be automatic for us, but not for our secondary pupils. If we want our students to automate complex concepts, we need to ensure sufficient time, focus, attention, revisiting, application, consolidation, practice, usage and eventual mastery. In many areas, I’m unconvinced we do this for them, as sixth form English students’ weak writing and history graduates’ weak knowledge testifies.

Why cramming fails; and why you retain rusty languages

Robert Bjork’s model of memory is a grid of storage strength and retrieval strength. Storage strength is how well learned something is. Retrieval strength is how accessible it is. I have redesigned it here:


This explains why cramming fails: the GCSE physics exam material I crammed into my mind had a high retrieval strength, as I could access it on exam day, but low storage strength, as I never learned it very well in the first place – and now I can’t remember it at all. It was crammed, and is now forgotten.

This also explains why though I’m rusty at French, which I spoke fluidly at 18, I could get it pick it up again quickly if I went back to France for a week or two; my French has high storage strength but low retrieval strength. It was mastered, is now buried, but can be remastered.

In either case, if I want to remember more of my Physics or more of my French, the route is the same: increased storage strength and increased retrieval strength is required for mastery.

So how can we help students remember what they’ve learned?

Willingham makes some suggestions:

1. Distributing practice (rather than cramming):it is virtually impossible to become proficient at any mental task without extended, dedicated practice distributed over time.’

2. Overlearning: keep pupils learning after they know the material to prevent forgetting: ‘a good rule of thumb is to put in another 20 percent of the time it took to master the material’.

3. Testing frequently: testing students frequently helps them remember material.


Bjork makes similar suggestions:

1. Spacing (rather than massing) practice: information that is presented repeatedly over spaced intervals is learned much better than information that is repeated without intervals.

2. Interleaving: although people think that they learn better when content is blocked, rather than interleaved, people actually learn content better when it is interleaved with other content.

3. Testing: using our memory improves our memory: the act of retrieval helps us remember the things we recall.  When information is successfully retrieved from memory, its representation in memory is changed such that it becomes more recallable in the future (Bjork, 1975); and this improvement is often greater than the benefit resulting from additional study (Roediger & Karpicke, 2006).


A great illustration of how counterintuitive the testing effect is comes from David Didau:

‘Which of these study patterns is more likely to result in long-term learning?

  1. study study study study – test
  2. study study study test – test
  3. study study test test – test
  4. study test test test – test

Most of us will pick 1. It just feels right, doesn’t it? Spaced repetitions of study are bound to result in better results, right? Wrong. The most successful pattern is in fact No. 4. Having just one study session, followed by three short testing sessions – and then a final assessment –  will out perform any other pattern. Who would have thought?’

How many of our school curricula and assessment systems are designed with these principles in mind: spacing distributed, interleaved practice; overlearning to mastery; frequent, low-stakes testing? How many teachers are aware of the research into working memory, long-term memory, storage strength and retrieval strength? How many of us in education apply these insights in our curriculum design or day-to-day teaching?


If this diagnosis is right – that the main reason pupils find it so hard to remember subject content is that our curricula and assessment aren’t designed with memory in mind – we are fortunate in one sense: there is a clear remedy. We must redesign our school curricula and assessment with memory in mind. That is the subject of my next post.

Bjork’s research can be read here:


Willingham’s research into memory can be read here:


“Practice Makes Perfect—But Only If You Practice Beyond the Point of Perfection,” American Educator, Spring 2004, http://www.aft. org/pubs-reports/american_educator/spring2004/cogsci.html

“Why Students Think They Understand—When They Don’t,” American Educator, Winter 2003-04, www.aft.org/pubs- reports/american_educator/winter03-04/cognitive.html.

“Students Remember … What They Think About,” American Educator, Summer 2003, http://www.aft.org/pubs-reports/ american_educator/summer2003/cogsci.html.

“Allocating Student Study Time: ‘Massed’ versus ‘Distributed’ Practice,” American Educator, Summer 2002, http://www.aft.org/ pubs-reports/american_educator/summer2002/askcognitive scientist.html.

A great summary by David Fawcett on the implications of the research for planning can be read here:



A great example of a teacher applying this research and how a story mnemonic can help students remember a mathematical formula can be read on Kris Boulton’s blog here:


About Joe Kirby

School leader, education writer, Director of Education and co-founder, Athena Learning Trust, Deputy head and co-founder, Michaela Community School, English teacher
This entry was posted in Education. Bookmark the permalink.

46 Responses to Why don’t students remember what they’ve learned?

  1. Pedro says:

    Dit is op From experience to meaning… herblogden reageerde:
    Great overview with a lot of practical insights on learning and trying not to forget!

  2. Eddie Carron says:

    I am a little skeptical about the role of ‘repetition’ in memory. At age 84 I remember a great many things that happened to me only once – and very briefly. I think ‘intellectual engagement’ is the key to memory – if we fail to engage a pupil’s intellect – if his/her thinking processes remain alternatively engaged, there will be no memory of the target idea.

    I was ‘told’ to learn ‘The Charge of the Light Brigade’ for homework about 72 years ago and naturally, i didnt bother until 30 mins before the English lesson where I read through it breaktime. We ‘did ‘ the poem once only but I can remember every word in every line to this day.

    I am involved with securing literacy skills among non or near-non readers – I know the key to becoming a competent, skilled reader is ‘lots of reading’ but I know that the reading material has to engage the intellect of the readers or it will fail. Within this ‘lots of reading experience’ children who had been failed by the phonics treadmill, intuitively assimilate the phonics correspondences without them being ritually ‘taught’. The children learned effortlessly, that which ritual ‘teaching’ had failed to secure proving that it was the ‘teaching’ that had failed and not the children who ultimately pay the price of our failure.

    I think that a good ‘professional’ understanding of ‘teaching ‘ is needed in order to improve learning.

  3. Pingback: Why don’t students remember what they’ve learned? | Pragmatic Education | Learning Curve

  4. Pingback: Why don’t students remember what they’ve learned? | LeAnne Hitt

  5. Michael Tidd says:

    ‘a good rule of thumb is to put in another 20 percent of the time it took to master the material’

    Such an important message, but it’d be a brave teacher who used a lesson during Ofsted to apply it!

  6. Pingback: Why students don’t remember what they have been taught | Phonic Books

  7. Thank you so much for this Joe. I’ll be using this with PGCE History trainees. We already do quite a lot on memory but this crisp and provocative summary will add much to our work.
    Long before I had the sense to look for research on this, I remember, as a Head of History at my first comprehensive school, theorising a simplified model, just on the basis of having tried to work out how retention seemed to occur and what one needed to do for those pupils who seemed neither to retain nor to see the point of putting effort into retention. I began to think in terms of ‘fingertip’ and ‘residue’ knowledge. These are not quite proxies for working and long-term memory but they bear relationship with that model. ‘Fingertip’ was a useful idea because it suggests absolute thoroughness. I really do have to work to make sure I have Elizabeth’s parliaments or 19th century Reform Acts at my ‘fingertips’. In other words, I know when I have them with such facility that I can ‘move about’ within my knowledge in my own head, calling them up, deploying them, eventually arguing with them. Only if it once became fingertip is it likely to gain lasting memory presence at least as prototypes that can become residue (the residue that stays in the sieve… sorry, mixed metaphor). In other words, I may temporarily forget the really minute moves, the things particular figures said in Parliament or the month in which something occurred (the high-resolution elements) but sets of structures at mid-resolution will only stay precisely because they once WERE once high-resolution. In other words, I cannot miss out high-resolution, even if mid-resolution is my goal. High-resolution matters, both in its own right and by extension and association.
    Gradually, in this crude way, I worked out that if I got the lower-attaining and struggling pupils to invest in committing stories, events, details to memory (I used lots of low-stakes testing, with a range of methods, some planned, some as ‘surprises’), with great thoroughness – really working til they could confidently work with the material in their heads – they were able, six months later (and I cannot claim more than this) at least to use sets of prototypes sensibly. By prototypes I mean vital mid-res knowledge the absence of which typically mars all lower-attainers’ work e.g. not muddling up Parliament and government; remembering structure and functions of institutions, remembering outline chronology accurately; able to recall more abstract nouns and their specific period instantiations, confidence with all names and key events). This also acted as a much stronger basis for them to assimilate new material in other periods. That was the most striking thing… the patterning of ‘resonance’ that means a child apprehends the word ‘revolution’ or ‘franchise’ with entirely different ears from those that they had before they had encountered that word in another narrative.
    And I discovered that you have to work, work, work at this, but if you build up a classroom culture where it is normal, it is liberating. Far from being some horrible oppression, a culture of low-stakes testing is energising and can be spliced in, efficiently with other kinds of complementary historical learning (‘interleaving’ is interesting here).
    When I began training history teachers, I first reduced this just to two principles: ‘clarity’ and ‘thoroughness’ (one group of PGCE trainees bought me a T-shirt with the two words on front and back). But I later decided that there is more to this than being ‘clear’ and ‘thorough’, one has to theorise the different types of relationships one kind of knowledge has to another, and to link this to a host of disciplinary and language considerations.
    Thanks for continuing to drive this forwards.

  8. bt0558 says:

    Great summary and explanations.

    I am not a ‘core knowledge’ believer, but I do believe that knowledge is the foundation for understanding and problem solving.Key facts, key terms and key concepts are the basis of everything I do and I am looking forward to the next blogs in the series after which hopefully I will do things a bit better.

    Thank you

  9. Neil Gilbride says:

    I am uncomfortable that the suggestions for making curricula is based on one model of memory. I think (i’ll check this) at the last count there were 16 models of memory! It would have been interesting to see you compare other models, particularly which look at the interactions between working memory and long term and the implications for practice.

    One in particular would be baddeley and hitch’ breakdown of the working memory into Visio spatial scratchpad, phonological loop and central executive. I would recommend reading these as they have vast implications for what you are saying and taking to practice – ie you is saying that attention is driven outside of memory, whereas B and H model says the CE is THE control system. Plus, knowing this could help teachers think how they are engaging with rehersal and how these separate modules within the working memory can be “tapped”. Like in this model, the working memory is crucial, b

    Also, the position of memory and its interaction with other cognitive systems is in debate and again has implications for practice. information processing models of memory suggest that memory isn’t the top of a hierarchal system which is a central pillar running out to other skills such as reasoning. Information processing models of memory would suggest that Reasoning, comprehension and learning are intertwined with working memory as a joint partner. Thus to develop these skills would also develop the individuals memory (Case 1992, Halford 1993).

    Its great to see more cognitive focused work on your blog, but don’t fall in the trap of the (Ps love his book!) Lemov quoting “brain science”…(vomit).. just because cognition is further ahead then neuroscience doesn’t mean its not a minefield!

  10. Nick Hart says:

    Reblogged this on Penn Wood CPD.

  11. Pingback: » very interesting. RT @shoshyart: http://t.co/w8BXM…

  12. Reblogged this on paddington teaching and learning and commented:
    There’s lots of important research on the cognitive science of memory in the blogosphere at the moment. Here’s a fantastic post from Joe Kirby, with links at the end to some really interesting sources.

  13. dodiscimus says:

    I agree with all the suggestions about how to improve the amount that pupils remember but it can be hard to get the right balance between developing more advanced ideas and mastering the basics. Essentially this is a differentiation issue; it’s often the case that some pupils are really solid on a basic idea and ready to apply it at a higher level, whilst other pupils have got the idea well-enough to do some higher level work with it but actually need more practice if they are to retain the basic stuff. It’s very tempting to go to the higher level while you can, and noticeably holding back the more advanced students if you don’t. The solution is some effective differentiation but, as always, that’s easier said than done.

  14. Pingback: 365 Days in my shoes Day 325 | high heels and high notes

  15. Pingback: Life after levels: where SLT fear to tread | Pragmatic Education

  16. Pingback: What I learned from learning the periodic table and other thoughts on memory and retention of historical knowledge and understanding | Improving Teaching

  17. Pingback: Life after levels: who’ll create a mastery assessment system? | Pragmatic Education

  18. Pingback: Why we shouldn’t close down the skills-knowledge debate | Pragmatic Education

  19. Pingback: How best to teach: knowledge-led or skills-led lessons? | Pragmatic Education

  20. Pingback: How to plan a knowledge unit in English | Pragmatic Education

  21. Pingback: Books, bloggers & metablogs: The Blogosphere in 2013 | Pragmatic Education

  22. Josie M says:

    Joe, thanks for this post. I find memory – particularly in relation to education – a fascinating subject, since much of my role as Literacy Leader (Secondary) involves teaching students to read and requires them to come back regularly to build on their skills. I’m an advocate of the relearning and overlearning principles you mention, in order to move information from the short-term to the long-term memory. For our weakest Y7’s, we even intentionally schedule sessions at the same time each day, for the importance of routine.

    I agree with Eddie’s comment above (first in list) that for something to become a lasting memory, it requires more than repetition, but a necessary active engagement of the brain in whatever the topic / skill / concept is that is being taught. However, I’m pretty certain we’ll never crack the long-term memory issue if we expect students to merely engage with a topic and then remember it again months down the line, having only visited them once or even twice…

  23. Pingback: The cheat codes to intelligence: touchpaper#7 | Pragmatic Education

  24. Pingback: Free Thinking: I agree with Katharine | Pragmatic Education

  25. Pingback: ResearchEd: cognitive science in the classroom | Pragmatic Education

  26. Pingback: A guide to this blog | Pragmatic Education

  27. Pingback: One scientific insight for curriculum design | Pragmatic Education

  28. Pingback: One scientific insight for curriculum design | Teachers Blog

  29. BB2 says:

    Reblogged this on The BB2 Collaborative.

  30. Pingback: The Benefits of Regular Retrieval Practice | Class Teaching

  31. Pingback: Socrative for frequent, low-stakes testing | mrhousemanblog

  32. Pingback: Even Better If…I add quizzes | #TeachGeog

  33. Pingback: How can I simply implement cognitive science findings to improve learning? | EviEd

  34. Pingback: Curriculum Design 3: A common vocabulary, or a common misunderstanding? | Ramblings of a Teacher

  35. Pingback: Resources to support our research Triads, curated by Richard D | UCL Academy

  36. Suleiman says:

    to read read and read it realy help the slow learners i think it will be good for any students

  37. Pingback: Reflections on learning Chinese 2: how I memorised characters | mathagogy

  38. Pingback: Blogs of the Week – 11 December, 2015 | Rhyddings Learning Power

  39. Pingback: My journey into education | Mayan's Blog

  40. Pingback: The Blogosphere in 2016: Roaring Tigers, Hidden Dragons | Pragmatic Education

  41. Pingback: JUSTIFYING YOUR PEDAGOGY – paul g moss

  42. Pingback: What is ENGAGEMENT? – paul g moss

  43. Pingback: Treasure Trove #4: Cognitive Science Crash Course | Joe Kirby

  44. Pingback: Treasure Trove #5: A Cognitive Science Crash Course | Joe Kirby

  45. Diane Olatutu Odelola says:

    I am a Head of a primary school. My pupils were doing poorly in maths and they had exams to write as they are in primary 6.

    We could not figure out what was responsible for this poor performance. In desperation I turned to God first and then to Google!😂 where I found your article on why pupils forget what we teach them! I know a little about short and long term memory, but your article opened a whole world of more detailed information to me which I am going to share with my staff. You are absolutely right- all teachers need in-depth understanding of cognitive science.

Leave a Reply