What is retrieval practice?

Have you ever tried to remember the name of a film or a song or a restaurant? Or perhaps an acquaintance’s name? Or your spouse’s? You know it, you do, but it is a real struggle to recall it. It’s right on the tip of your tongue. Well, that struggle, as long as it’s repeated, is great for remembering things. That, in a neurological nutshell, is retrieval practice.

The ‘retrieval’ bit of retrieval practice means recalling something from long-term memory and putting it into working memory. The ‘practice’ bit of retrieval practice means doing it more than once.

What’s long-term and working memory?

Your long-term memory stores, well, your long-term memories. Functionally, it’s limitless in size and (in theory at least) can store memories permanently. Your working memory is the bit of the brain where thinking takes place. Unlike your long-term memory, your working memory can’t hold much information (about four items) and it can’t hold it for very long either. Once you put your attention somewhere else, about half a minute later, what was in your working memory will go up to the neurological ether and, poof, disappear.

How does retrieval practice work?

With retrieval practice, you are retrieving things from your long-term memory and putting them into your working memory. Through that process, you lay down memory traces – neurological way markers, if you like – which help you find that information again in the future.

Think of a walk through a dense forest. The first time you go through that forest, it’s difficult to find your way. It’s overgrown. Everything is unfamiliar. You quickly lose your bearings. But every time you take that route again, the route becomes increasingly familiar. You recognise that fallen tree, that odd-shaped branch, that tiny stream with the wobbly stone. Consequently, navigating your way through becomes easier and easier – to the point, in fact, that you don’t navigate at all, you just do it. It becomes an unthinking act.

So it is with retrieval practice. Every time you retrieve a memory, the pathway to that memory becomes more established, more definite, more, so to speak, neurologically well-trodden. Consequently, finding that memory becomes easier and easier – to the point, in fact, that there’s no more struggle: the memory is just there for you to access anytime that you need it. As Professor Robert Bjork (Belluck 2011) succinctly puts it: “What we recall becomes more recallable.”

There was a time in your life when you couldn’t do up your shoelaces. In fact, you might still have a faint memory of when you first tried to learn this rather vital life skill. It was hard work. It required concentration. You had to really think about it. Makes sense, it’s tricky thing, learning to do up your shoelaces. The process is staged, sequential and fiddly. But now, presumably, you can do up your shoelaces easily, effortless and even unthinkingly (at least without conscious thought). Why? Because the process has been encoded into your long-term memory.

Your long-term memory stores, well, your long-term memories. Functionally, it’s limitless in size and (in theory at least) can store memories permanently. Your working memory is the bit of the brain where thinking takes place. Unlike your long-term memory, your working memory can’t hold much information (about four items) and it can’t hold it for very long either. Once you put your attention somewhere else, about half a minute later, what was in your working memory will go up to the neurological ether and, poof, disappear.

Same goes for knowing your times tables, speaking a foreign language, speaking a native language, knowing that the capital of France is Paris (and F), knowing what happened in 1066 and where it happened, touch typing, touch texting, driving, cycling, walking, and looking right when you cross the road. Regarding that last example, have you ever stepped into a road after checking over your right shoulder to see if it was safe to do so, and still got beeped at? If you have, it’s called walking in mainland Europe (or any other area of the world where they drive on the other side). So firmly encoded is the requirement for us to look right, it becomes an automatic act – to the point that it can actually put us into a bit of mortal danger. It’s why I always spend my two weeks in Budleigh Salterton.

What’s the evidence for retrieval practice?

Lots and lots and lots, going all the way back to Edwina Abbot (1909) and, before that, the related work of Hermann Ebbinghaus (1881) and his now-famous (but should still be more famous) forgetting curve. In fact, as an idea the importance of retrieval practice goes back millennia:

Exercise in repeatedly recalling a thing strengthens the memory.

Aristotle 384 – 322 BCE

A seminal modern-day study is Karpicke & Roediger (2008). They compared retrieval practice with studying (aka reading and re-reading). The participants (college students) had to learn 40 Swahili words and the corresponding English translation (e.g. mashua / boat). It was all done via a computer screen. There were two groups and both groups started by studying (reading) the Swahili words and the English translations. Group one then continued this approach, whereas group two were repeatedly tested – the Swahili word would pop up on the screen and they had to type the English translation. The groups came back a week later to see how much they could remember. Group two, the group that had been repeatedly tested, had double the recall of group one, the study group.

So, retrieval practice works and restudying doesn’t?

Not quite. Both work but retrieval practice is much more effective. With retrieval practice you’re not only encoding the information, you’re also encoding how to find that information again. Journey and destination, if you like. With reading and re-reading, it’s just the destination that’s being encoded, not how to get there.

The most common way that students revise is by reading and re-reading their notes (or textbook, study guide, knowledge organiser etc). We know it’s not as effective as retrieval practice, not by a long way, but it feels right. After all, there’s a good chance that to understand the material in the first place, they had to read about it, and given that they now understand it, clearly reading is effective. So, it makes sense to think that re-reading would be effective too. But it isn’t.

Not only that, re-reading is easy. Unlike retrieval practice, it doesn’t hurt your head. It’s effortless, not effortful. You’re not being reminded of how much you don’t know, not being faced with the gaps in your knowledge. Actually, quite the opposite: everything you read is reassuringly familiar. And that familiarity creates the illusion of learning.

Moreover, re-reading does actually work. It does. It encodes information into your long-term memory. It’s just that it doesn’t do it very well. Not nearly as well as retrieval practice. With retrieval practice you remember more; you remember it more quickly; and you remember it more durably.

So, when it comes to remembering stuff, retrieval practice is the best approach, yes?

Nearly. Retrieval practice is a very powerful strategy. However, it is made even more powerful when it’s spaced out over time – something called, unsurprisingly, spaced retrieval.

Karpicke & Bauernschmidt (2011), using the Swahili / English word approach that Karpicke & Roediger (2008) employed in their seminal work, compared the following four conditions:

  • a single period of study (‘study once’)
  • a single retrieval (‘retrieve once’)
  • three retrievals in a row (‘massed retrieval’)
  • three retrievals spaced over time (‘spaced retrieval’)

As with the 2008 work, the participants came back a week later and were tested to see what they could remember.

Bar chart adapted from here.

A couple of things stand out from this experiment. First, the ineffectiveness of studying once. Second, that massed retrieval (aka cramming) is no better than a single retrieval, at least when the measure is week-long retention. There is evidence that says that cramming a night before a test is useful, but the material won’t be durable. And third, that spaced retrieval is by far the best approach.

However, there’s no clear consensus about what constitutes the most effective retrieval schedule. Certainly, given that we forget over 50% of information within an hour, the first act of retrieval should happen pretty much immediately. Makes sense. But after that, it’s more difficult to say and probably depends on what needs to be encoded. It’s a fair assumption, though, that the more complex the material, the more frequent the scheduling needs to be. And the following seems a reasonable starting point:

hour → day → week → fortnight → month → two months → six months (and so on)

What is known for sure is that the intervals (‘spaces’) between the retrieval practices should increase over time. These intervals are important because they lead to a degradation of memory traces – the neurological route to how to find that memory again. It’s a paradox, but the act of forgetting seems to help the act of remembering.

It appears to work like this: the act of forgetting makes retrieval practice more difficult. This increased difficultly strengthens the neurological pathway. The strengthened pathway makes it easier to find the memory again. It’s not clear what’s going on neurologically in this process, but it may well be that memory traces are being laid down (as it were) on top of memory traces, and because there’s difficulty involved, the brain devotes extra resources (nutrients etc) to this process. But frankly it’s not clear.

That’s interesting. So is difficulty desirable then?  

Yes. Exactly. Difficulty is very desirable (to slightly paraphrase Robert A. Bjork).

Now, of course, retrieval practice gets easier over time. That’s the point of it. But we want to keep the difficulty in place. Spaced retrieval (the intervals) is one way of doing that. Other ways include:

  • adding extra items
  • incorporating time restraints
  • increasing complexity
  • upping the success criteria

To give an example. Let’s say someone wants to learn all the countries in Africa, names and locations. They can begin by retrieving, say, ten of those countries. The number then goes up in increments of five. The time they are given to retrieve the countries decreases over time – or, if extra countries are added, the time doesn’t increase. Once all the countries have been successfully retrieved, then they list all the coastal countries in order starting at Morocco and working their way around the continent clockwise. And then anti-clockwise. Or they have to list the countries in order of size. Or alphabetical order. Or reverse alphabetical order. Or … you get the idea. The success criteria for all activities is initially set at 80% but moves up to 90% and then 100%. And once they have done all of that, they move on to remembering all the countries in Europe alongside those African countries. And then all the countries in Asia and Europe, alongside those African countries. And so on.

This desirable difficulty idea relates to something called the retrieval effort hypothesis. Simply put, it states that “difficult but successful retrievals are far better for memory than easier successful retrievals” (Pyc & Rawson 2009). To put it another way, if the difficulty isn’t there, then you get diminishing returns on your neurological investments. So make it difficult – but make it achievably difficult. If it’s impossible the students won’t engage with it. It has to be within their reach, as long as that reach includes a bit of a stretch and they are on tiptoes. With this approach the impossible, over time, becomes possible.    

Ok, I’m convinced. Retrieval practice is good. So, how do you do it?

Glad you’re on board. Here’s a ready-made list of things you can do. You’ll be familiar with most of them, but don’t let that familiarity fool you into thinking they’re not powerful, because they are. Very. Here’s the list:

1. Questions

A particularly effective question format is something called a PPPB. It stands for pose, pause, pounce and bounce. You POSE a question. You PAUSE for thinking time. That’s important because students need time to retrieve the information. On average, teachers give 0.7 seconds thinking time when they ask a question – if a student can answer that fast, that’s not retrieval, that’s automaticity. Not only that, once the correct answer has been said out loud, then that’s the end of the retrieval opportunity for all the other students. So, don’t be afraid to wait. Ten, twenty, thirty seconds – or longer if the complexity of the question requires it. Ignore any hands that shoot up. Give the students time to think.

Then you POUNCE – metaphorically, of course. In other words, you choose a student to answer. It doesn’t matter if the student is volunteering or not, because you’ve created a culture in your classroom where every student is expected to answer. It’s called ‘cold calling’ and it’s great for increasing student engagement. You see, because any student can be asked, it makes sense that every student thinks about the answer, just in case it’s them.

Some teachers don’t like cold calling. Their worry is that it will have a negative impact on the student’s self-esteem should the answer be incorrect. But hold on: we want high challenge and low or no stakes. It doesn’t matter if the student gets it wrong. In fact, it is completely fine if they do because getting something wrong is a necessary stage in the process of getting something right. As the great Yoda himself proclaimed, “The greatest teacher, failure is.”

Anyway, even if the student does get it wrong, that’s good because it leverages in something called the hypercorrection effect (Metcalfe 2016). In other words, a student will remember the answer better if they get it wrong than if they guessed it right in the first place – particularly if they were confident in their incorrect answer.

So, what do you do if they get it wrong? Well, you have choices. You could simply give the answer. That’s effective. Or offer some clues so that the answer can be found. Or you can go back a few steps and take the student along the journey to the right answer. Or you can BOUNCE the question to another student to answer. Once the correct answer is given, you then BOUNCE it back to the first student so that he or she can repeat the answer. And praise them for doing so.

In his fine book, Why Don’t Students Like School (2009) Daniel Willingham says that ‘memory is the residue of thought.’ It makes sense, then, that when we ask students questions, we want all of them thinking, all of them retrieving, all of them laying down those memory traces. PPPB makes this more likely.

2. Tests

Retrieval practice is based on something called the testing effect. Keep on testing yourself and retrieval improves. It’s nothing more complicated than that.

Now, tests have a reputation for high challenge (good) but also high stakes (bad). The very idea of a test can bring people out in a cold sweat. It triggers ideas of assessment and judgement – maybe even a wagging finger or a disappointing look. The night before a test can be sleepless; the hours after a test, fraught with worries of how well you’ve done. For many, a ‘test’ is the second nastiest word in the English language – the only one that’s nastier being ‘exam’.

So it’s important that students fully understand that when tests are being used as retrieval practice, there is no assessment or judgement. Yes, the tests are challenging, but they are like that to help them learn, not to access how much they’ve learnt. The goal is low stakes or no stakes. Framed that way, the students are more likely to feel free to get things wrong, and that’s vitally important because making mistakes is a fundamental part of the learning process.

One other thing: let’s not forget that tests are a reality of formal education. Summative tests wait for students at the end of every key stage. So, having lots of testing in your lessons (including questions from past papers) will help students get use to taking tests. Which means that when they get to their SATs or GCSEs or A Level exams, their stress levels will be much more manageable. This point links directly to memory because when we feel stressed, our memory is adversely affected.

3. Quizzes

Of course, a quick way to sidestep the negative connotations of a ‘test’ is not to call a test a test, but instead to call it a quiz. Students love quizzes. So have lots and lots of quizzes. In fact, make quizzes a part of your daily routine. Frame them positively and inject a bit of fun: “Ok, you lucky, lucky students, it’s time for our morning quiz. What a wonderful way to start the day!”

Include questions on what you did last lesson (that’s very important), but also include questions that loop back to previous lessons and topics. A strategy called ‘sevens’ works well. As you can probably guess, the strategy consists of seven questions. Questions 1, 2 and 3 are about what was covered last lesson; question 4, last week; question 5, last month; question 6, two months ago; and question 7, six months ago. Yep, it’s spaced retrieval. Of course, you can switch to ‘threes’ or ‘fives’ or ‘nines’ or …  you get the idea. But whatever you do, have lots of quizzes, have them daily, and keep on looping back to earlier material.

Students can also create their own quizzes. They can self-test or get tested by their peers or parents. Quizzes can be created from their notes, from textbooks, study guides, knowledge organisers – the only criterion is that the original source is factually correct. Students can take a photo of their quiz and send it to their friends to do, and their friends can return the favour. Another photo showing the answers should also be sent – you always need the answers so you can see what it was you remembered and what it was you forgot.

There’s also another reason why you need the answers. We know that people learn from answering quizzes and tests. Now, if that person has given the wrong answer and it hasn’t been corrected, then there’s a good chance that they will encode that wrong answer – in fact, it will be more encoded than if the question hadn’t been asked in the first place. This is true for all types of questions, but particularly so for formats that set up an error option, like true/false questions and multiple choice. If the error option is chosen and it’s not corrected, the error gets stamped into long-term memory. But if it’s corrected right away, then this negative suggestion effect is no longer a problem.

4. Flashcards

Flashcards, as you know, have a question on one side and the answer on the other. Flashcards are either fact-based or concept-based:

Fact: when did Bucks Fizz win the Eurovision Song Contest?
Concept: music – vocal or instrumental sounds (or both) combined in such a way as to produce beauty of form, harmony and expression of emotion

Students can make the cards themselves or be given the cards. While there is some small encoding benefit to the former, it is so outweighed by the actual use of flashcards as a retrieval activity that it doesn’t really matter if they don’t make them themselves. But, either way, they can add more cards as the weeks go by. In fact, some teachers set aside time every week to do just that. The information on the flashcards can be words or pictures or both (don’t forget the power of dual coding – when you have the same information as both words and images, you have two ways of remembering that information later on).

A word of warning. Flashcards, like all retrieval activities, can hurt your brain a bit. They require mental energy. Finding the answer can be a real strain. And, of course, the answer is on the back of the card. Consequently, students might be tempted to turn over the card too soon. In fact, it turns out that that’s exactly what most (about 70%) do – see here). But they must not. Do that and they’ve retrieved nothing. So you need a rule, either ‘say it before you see it’ or ‘write it before you read it’. If they can’t do that, then they need to put the card to one side and return to it once they’ve gone through all the other cards. If they still can’t do it, then – and only then – can they turn over and take a look. Adding in this delay means that the students are more invested in finding out what the answer is. Which means, in turn, that the information will be more firmly encoded.

Another danger is that students stop testing themselves on the flashcard once they think they know it. And, of course, at that very moment, they do know it. But knowing something is not the same as having that thing encoded into long-term memory – let’s not forget Ebbinghaus’s forgetting curve (pun intended) and the rapidity of forgetfulness. So, students must keep on returning to those flashcards. The six-box strategy can help with this.

Students get six boxes, numbered and labelled like this:

Students put all their flashcards in box 1. They then take out their flashcards and work through them. If they get a card right, it goes into box 2. If they get it wrong, it goes back into box 1. The next day they take out all the cards from box 1 (the everyday box) and go through them. If they get a card right, it goes into box 2, if they get it wrong, it goes back into box 1. Then (because it is still the ‘next’ day) they go through the flashcards in box 2. If they get a card right, it goes into box 3, if they get it wrong, it goes back into box 1. Hence, every time a card is answered correctly, it graduates to the next box; every time it is answered incorrectly, it goes back to box 1, regardless of where it was in the sequence. This, of course, is an example of spaced retrieval.

Flashcards are great for learning factual information. But they can also be used for more conceptual information. Here’s how. Students make a pile of all the flashcards that are concept-based. They then make another pile of cards called ‘instruction cards’.

The instruction cards instruct the students to engage with the concept cards in a particular way. Here are some examples, many of them coming from this blog post:

  • Describe a film/television programme or scene that depicts the concept
  • Draw this concept (in one minute or less)
  • Give a real-life example of this concept
  • Pretend to explain this concept to a 5-year-old child
  • What is the opposite of this concept?
  • How is this concept connected to your life?
  • What would happen to the world if this concept was removed?
  • Why is knowledge of this concept useful to you?
  • Give an argument why everyone in the world should know this concept
  • Which groups of people might this concept be most important to?
  • Come up with a new name (neologism) for this concept
  • If you met someone who didn’t believe in this concept, what would you say to them to convince them that it is true?

One instruction card per concept card works well. Next time you do the activity, shuffle to get a different combination of cards.

One last thing: flashcards make great question cards for quizzes (the previous strategy) and games (the next strategy).

5. Games

There’s no reason why games shouldn’t be used for retrieval practice. There are lots of familiar formats out there that can be used in this way. For instance:

Taboo – in this game a student has to guess a fact or concept. However, the person giving the clues is prohibited from using certain words – hence the game’s name, Taboo. Flashcards are perfect for this. The definition on the reverse of the flashcard becomes the taboo words (grammatical words like ‘and’ and ‘the’ are permitted). The game is difficult and fun in equal measure.

Just a Minute – like the Radio 4 game, students have to speak for one minute without hesitation, repetition or deviation on a given topic. A fourth criterion is that what they say must be factually correct. If the student fails in any of those areas, another student can challenge and then they take over and carry on speaking for the remainder of the minute.

Bingo – the bingo card is a sheet with boxes (3×3, 4×4, 5×5). The students number the boxes themselves (small, in the corner) so that each sheet is random. The teacher asks the questions and at the end of the game goes through the answers. The team that gets the first straight line across the whole card (horizontal, vertical or diagonal) wins. The second winner is the team that gets most answers right.

Snakes & Ladders – to go up the ladder the student has to get a question right, to stop going down the snake they have to get three questions right.

Pictionary – students draw the information and other students have to guess it. Pictionary also has dual coding benefits.

The Generation Game – students are shown items on a ‘conveyor belt’ (which in a classroom can be images or information on a PowerPoint slide). They then have 45 seconds to remember as many of those items as possible. For nostalgia sake, throw in an image of a cuddly toy.

One last thing about games. As with all activities, you want a high ratio of student engagement, so be careful about team sizes. Some of the games work best as a whole class activity but with students working on their own (e.g. The Generation Game). Some work best as pairs (e.g. Snakes & Ladders). And some work best as small groups (e.g. Just a Minute). As a rule, the smaller the team size, the better because that means there’s more opportunity for everyone to get involved. For Pictionary, you can have small groups or you can divide the class into two big teams.

6. Lists

On a blank piece of paper students list everything that they can remember about a lesson or a topic in, say, 3 minutes. They then compare their list with their partner’s list to see what they remembered and what they forgot. Pairs can then share with a second pair or the students can refer to their own notes or source material. Or the teacher goes through the list. You can start your lesson with this activity, and then repeat the activity at the end of the lesson, for yet more retrieval practice.

The list activity can be scaffolded if necessary (e.g. clues and trigger questions); over time, the scaffold can be removed to increase challenge.

Lists can be in list form (a column of information) or they can be incorporated into a graphic organiser.

7. Graphic organisers

A graphic organiser is way of organising information graphically, that is, using words and images to show relationships. So, for instance, instead of getting the students to list, say, the factors that led to the start of the First World War, you get them to retrieve that information and put it into a timeline:

Other graphic organisers include mind maps, cause and effect sequences, continuums, cross-continuums and Venn-diagrams. Graphic organisers can also incorporate arrows, bracketing devices, colours, diagrams and line drawings. By using graphic organisers, students are not only retrieving information, they are also finding connections and relationships within the information. So both retention and understanding are enhanced.

You can pick up a fantastic and free graphic organiser poster here (and lots of other great resources), all from the wonderful Oliver Caviglioli.

8. Knowledge organisers

A knowledge organiser sets out the key facts and information about a topic on a single page, typically A4 or A3. Here’s one on apartheid in South Africa:

A knowledge organiser is not something that students create or do, but rather something that is given to them by the teacher, ideally right at the beginning of the topic. While it’s not a retrieval activity in itself, it’s a great resource for retrieval because it’s got all the key details. Hence, it’s great for tests – self-tests, or tests from peers, parents or the teacher.

Knowledge organisers can also be used with the next retrieval activity: look, cover, write, check.

9. Look, cover, write, check

This strategy has been around for years. It’s frequently used to help students learn their spelings spelllings sppelings spellings. The student looks at the word (i.e. studies it); covers the word so they can’t see it; writes it down (i.e. retrieves it); and then checks to see if they’ve got it right. If they’ve got it wrong, they do it again until they’ve got it right.

It’s neat and simple. It’s also versatile because it can be used for all information, not just spellings, and that includes the information in a knowledge organiser. For example, the students study a section of the knowledge organiser. They then cover that section with a bit of paper. They then write down exactly – exactly! – what’s written under the paper. Finally, they check to see what they’ve remembered and what they’ve forgotten. Over time, the ‘look’ and ‘cover’ stages can be skipped, so the students just write and check.

10. Fill in the blanks

A cloze exercise is an example of ‘fill in the blanks’. Or, to put it another way, ‘A _________ exercise is an example of ‘fill in the blanks’.

It’s another neat, simple and versatile activity. Again, it can be used with knowledge organisers. For example, parts of a knowledge organiser are left blank (four examples below). The students fill in the blanks and then they check to see how well they’ve done.

 

Eventually – and just think for a moment about the power of this! – the students could be given an entirely blank knowledge organiser to fill in. Now, if they can do that, then they definitely know the material.

11. Think-pair-share

You ask the students a question. The students think about their answer. They then turn to their partner and explain their answer. For example:

Write down as many things as you can remember about Lady Macbeth from yesterday’s lesson. Support your answer with retrieved references from the text. Two minutes. As always, no talking. Start … now!

Writing helps the students to focus. It also helps you see who’s working and who’s not working. A time element is useful because it creates a sense of urgency, a need to get the work done. It also communicates that time in your lesson is a precious commodity that needs to be used to the full.

At the end of the allotted thinking time (and be an exact timekeeper) students share their responses. If only one answer is required (say, the students have to explain a mathematical process), don’t tell students who’ll be doing the explaining. That way, they’ll both engage with the initial retrieval and thinking.

One way to designate who’s who in the pairs is to use As and Bs. A better way, though, is an idea that Doug Lemov details in his wonderful book, Teach Like a Champion 2.0. One of the students will be sitting closer to the window than the wall. So you say, “window to wall … go” and the one sitting nearer to the window does the explaining. Or you say “wall to window” and the student sitting nearer to the wall does the explaining. Or wall to door. Or door to wall. You get the idea. The student not doing the explanation listens carefully to see if they can spot any missed information or steps.

12. Use an App

A great app for learning your times table is Times Table Rock Stars (often shortened to TT Rock Stars) – in fact, the UK’s very own Secondary of State for Education, The Right Honourable Damian Hinds MP, says his own three children use it. And my own nephew, Isaac, is also a regular user and, frankly, he’s embarrassingly good. Here’s a fabulous twitter clip of a Year 4 boy who’s not even the best in the school – WOW!

Ankiapp is a flashcard app. You can use it to make your own flashcards, or you can download one of 80+ million pre-made flashcards. It’s a great teacher and student resource, though it takes a while to get your head around what to do.

A completely free app is Name the Nations. My wife and I now know the name and location of every country in the world. And we’re not competitive about it, either. No, hasn’t caused any arguments. Not one.

Apps are fun. You always have them with you. They’re great for filling in a few spare minutes: back of the bus, waiting for a friend, during a boring programme on TV, and, um, on the loo. And they give you instant feedback. You either get it right or you get it wrong, and you find out right away. Apps also show you how well you’re doing over time – that’s good because increased mastery is motivational. They also tell you what you need to do to achieve the next level/cup/medal/ribbon/award/tick/star – and that’s also motivational.

They’re also good for another reason. The app, not the user, takes control of the scheduling. We often overestimate our own abilities. We think we know something when we don’t and that we’ll remember something when we won’t. Consequently, we tend not to test ourselves enough. The app doesn’t share our ego biases or metacognition deficits, so is much more objective (and consequently much more helpful) in its scheduling.

Ok, so I now know what I should do. Any strategies to avoid?

As this study shows, re-reading and highlighting have low utility. In other words, they’re a bit rubbish. Highlighting is fine if you want to find something again, but other than it’s not worth bothering with. In fact, it might even contribute to the illusion of learning. All those highlighted words give the impression that you’ve done an enormous amount of work, but learning takes a lot more effort than a quick stroke of a pen – even if that pen gives a beautiful shade of pink/yellow/purple/green/sky blue.

Any other strategies to avoid?

Yes, study groups. Now, I don’t have any research evidence for this one. This is personal anecdote, from my own experience as a teacher. The problem with study groups, as I see it, is that often they’ve got nothing to do with studying. Study groups are, in fact, social groups. As social groups, they work brilliantly, but as a way of getting information firmly encoded into long-term memory, they’re awful. If a student really, really, really, really, really wants to be a part of a study group, then the first thing that needs to happen is a name change. Stop calling them study groups, instead call them testing groups. At least then there’s a chance, slim maybe, that some retrieval practice will take place.

Is there anything else I need to know about retrieval practice? Any further tips?

Yes, these:

1. The first retrieval practice needs to happen as soon as possible. We forget things very quickly so begin the encoding process right away.

2. Retrieval practice is most powerful when it is spaced over time (spaced retrieval). So keep on finding opportunities for students to retrieve previously taught material.

3. Make sure that the student sees the correct answer after the retrieval activity. It is not enough for them to simply know whether they got it right or wrong

4. If the test has an error option (multiple choice, true/false questions) wrong answers may be viewed as being correct. So do point 3!

5. Retrieval practice gets easier over time. So keep adding in extra difficulty (e.g. more items, more complexity, higher success criteria, less time). But …

6. Don’t make it too difficult. Difficulty is desirable, but it has to be achievable difficulty. If it’s too hard the students won’t engage.

7. If a student makes a mistake during retrieval practice, good! But the mistake cannot stand. They need to try again until they get it right.

8. Tell the students that if retrieval practice hurts their heads, then that means they’re doing it right: ‘no mental pain, no mental gain’.

9. Aim for low stakes or no stakes and inject a bit of fun. Students need to feel free to make mistakes.

10. Draw the students’ attention to the progress they’re making – increased mastery is a great motivator.

How do I find out more about retrieval practice?

There’s an increasing amount of material on the internet – type ‘retrieval practice’ into Google and see what you come up with. A particularly good website is The Learning Scientists – aka Drs Yana Weinstein and Megan Sumeracki. They also have a very good podcast.

Is that it?

Well, there’s one other thing I wanted to say, but for the life of me I can’t remember what. Blast! I should have done some spaced retrieval. Practise what you preach, Robin, practise what you preach.

References

Abbott E. E. (1909). “On the analysis of the factors of recall in the learning process.” Psychological Monographs. 11, 159–177

Belluck, P. (2011). “Test-Taking Cements Knowledge Better Than Studying, Researchers Say.” Nytimes.com. January 21, 2011. Available here

Karpicke, J.D & Bauernschmidt, A (2011). “Spaced retrieval: Absolute spacing enhances learning regardless of relative spacing.” Journal of Experimental Psychology: Learning, Memory, and Cognition 37(5), 1250-1257

Karpicke, J.D & Roediger, H.L (2008). “The critical importance of retrieval for learning.” Science. 319 (5865): 966–968. Available here

Metcalfe, J (2016). “Learning from Errors.” Annual Review of Psychology, Vol. 68, pp. 465-489. Available here

 Pyc, M.A & Rawson, K.A (2009). “Testing the retrieval effort hypothesis: Does greater difficulty correctly recalling information lead to higher levels of memory?” Journal of Memory and Language. 60(4): 437–447. Available here

Willingham, D. T. (2009). “Why don’t students like school? A cognitive scientist answers questions about how the mind works and what it means for the classroom.” San Francisco, CA, US: Jossey-Bass.