Doug Lemov's field notes

Reflections on teaching, literacy, coaching, and practice.

10.03.17Retrieval Practice: A Teachers’ Definition and Video Examples

Retrieval Practice: Image via Oliver Caviglioli


We’ve been reading up a lot on retrieval practice lately.  Hopefully we’re not alone in that. From a cognitive science standpoint it’s absolutely central to improving learning.

You might recall  Daniel Willingham’s assertion about the importance of knowledge:

Data from the last thirty years lead to a conclusion that is not scientifically challengeable: thinking well requires knowing facts, and that’s true not simply because you need something to think about. The very processes that teachers care about most — critical thinking processes such as reasoning and problem solving — are intimately intertwined with factual knowledge that is in long-term memory (not just found in the environment).

There are two parts of Willingham’s assertion:

1) You can only think deeply and critically about what you know well—what you have a lot of knowledge about—and

2)  To aid thinking, that knowledge must be encoded in long-term memory. Retrieval practice is the tool that encodes knowledge in long term memory.

Having to recall something -especially having to struggle to recall it after a bit of forgetting has begun–strengthens the neural pathway in which it is encoded making it stronger and easier to recall.

Of the two ideas—1) knowledge is important so 2) you have to encode it in students’ long term memory–the latter is probably more dissonant and difficult for many teachers because the encoding process requires what they think of as lower order questions—retrieval practice. Even if you know students need knowledge and you know that simple questions now support rigorous questions later it can be hard to make yourself do.  It’s one of the most important things we can do to ensure learning and yet we call it or think of it in the back of our minds as ‘regurgitation.’

As part of our upcoming workshop on Ratio this month we’re going to talk about the connection between knowledge and higher order thinking and so we’ve just finished cutting our first video of teachers using retrieval practice in the classroom.  We’ve got six teachers two math, two history and French teacher and an English teacher in the montage.

Here it is.  We hope you can steal some useful pointers.

EA.RetrievalPractice.GRM.MultipleTeachers.’Retrieval Practice Montage.’Clip2729 from TLAC Blog on Vimeo.

One thing we thought of as we watched was that we needed a teachers working definition of ‘retrieval practice’-a more specific definition than the one we got from the ever-helpful Pooja Agarwal, one of the leading scientists studying retrieval practice. She defined retrieval practice as “remembering,” or “pulling information out instead of putting it in.”

For classroom use however we wanted something more specific so we’re going to use this:

Retrieval Practice occurs when learners recall and apply multiple examples of previously learned knowledge or skills after a period of forgetting.

Here are some key things we were trying to highlight in our definition:

  • Retrieval is relevant to skills (hitting a baseball, solving for perimeter of an octagon) or knowledge (dates in history; formulas, vocabulary).
  • In the classroom it involves groups of questions in blocks as opposed to arguing that I “ask retrieval questions all the time,” say.
  • It requires at least a short delay after something has been learned because once you’ve started to forget you have to work harder to remember and this creates a stronger neural pathway


Thoughts about the video:

Some of them are using ‘oral drill’ a form of the technique Pepper where students are asked to stand up, but the standing is irrelevant.  The key is that the teachers are asking students to recall facts and skills that they know.

Teachers are bringing retrieval practice to different parts of their lessons: Eric Snider begins his reading lesson with retrieval practice about the story. Annette Riffle uses retrieval practice as a high value activity for the class to do while one student is working at the board. Barry Smith builds large blocks of his lesson around it. Steve Chiger makes it a fast review at the beginning of class to get the energy level up.  Art Worrell uses it to review before a test.

The teachers use different approaches:  Rachael Taylor starts it as pair work in a Turn and Talk.  Jesse Rector and Art Worrell Cold Call.  Barry Smith uses a combination of taking hands, Call and Response and Cold Call.

Art asks students to build off each other’s ideas.  Barry goes back and forth French to English and English to French. Jesse asks two step problems.

But over and over we see learners recall and apply multiple examples of previously learned knowledge or skills after a period of forgetting.  The big misreading of the video would be to assume (or argue speciously) that what it implies is that these teachers’ classes look like this all the time.  What you are seeing is the moments when they stress retrieval practice and those moments look different from the other activities they do that rely on and benefit from the knowledge retrieval practice encodes.

Hope you find it helpful and look forward to hearing your thoughts.

Image result for retrieval practice

Retrieval Practice: Because Everybody Benefits


3 Responses to “Retrieval Practice: A Teachers’ Definition and Video Examples”

  1. Brian
    November 10, 2017 at 8:21 am

    I would like to make a couple of comments on this blogpost, as a teacher who has a keen interest in memory and it’s importance (as one of a number of factors) in cognition.

    From my perspective, you take several basic and easily understood ideas and conflate them to create some type of model that may be used to “improve” learning.

    I note the you do not quote Willingham directly, but appear to summarise his views on the topic. It is obvious that each and every day, people gather information which they combine with information from long term memory to solve problems and in the process may commit additional knowledge to long term memory. Having information available to solve problems may come from long term memory but for most of us I would question the extent to which this is necessary. I believe that the extent to which “knowledge” which you seem to wish to define as information held in long term memory is an interesting one and open perhaps to debate. Almost exclusively, when I need to solve a problem I access knowledge held by others (collaboration/teamwork) including “googling it” and I am able to provide informed and valid/sound arguments in support of my solutions and evaluations.

    You only talk about recall here and not recognition. Although it is nice to be able to recall information, spending less time on strengthening storage and using recognition approaches may be much more efficient and effective in the long run. “May”.

    You talk of “encoding” requiring “lower order questions” and in my experience there is no evidence that this is the case in fact I see the opposite. Indeed you also see some sort of requirement to “apply” recalled lnowledge to “strengthen pathways”. Application of knowledge often is and often requires very sophisticated cognitive processes which require analysis, evaluation and synthesising skills. I believe Willingham sees storgae of memories as a process which combines knowledge acquisition with higher order and more complex thinking skills, a view which I believe he conformed to me in a twitter discussion a while back.

    Clearly it would be silly of me to argue with the idea that retrieval of knowledge is a useful activity and that in some way(s) greater breadth and depth of knowledge will likely increase the speed and quality of problem solving. I would question the necessity for the former and suggest that the former has a significant negative impact on the latter more than any increases in fluency will have a positive impact.

    I believe you have fallen foul of your own soundbites, soundbites you have created to provide general advice which promises to provide a magic bullt to the issues considered.

    I suspect that in an attempt to commit as many “facts” to long term memory, you have in fact summarised to an unreasonable extent, you have generalised unreasonably and extrapolated beyond your data. To risk the ire of the “cognitive load advocates” I would humbly suggest that spending too much time of remembering facts, that is instantiations of conceptual ideas, teachers may end up reducing the student’s ability to autididactically. I believe that teaching should focus on conceptual understandings with sufficient factual knowledge to allow learners to ditch the teacher and embark on their own lifelong learning journeys. For me it is this balance that is important.

    To suggest that a key focus should be on maximising long term recall of factual information is I feel a potentially dangerous message unless an individual has an extremely effective memory storage apparatus. Unfortunately for many this is not the case and therefore your recommendations seem to me to favour good memory.

    If commiting facts to long term memory was the key to learning and functioning as a human being, I would have thought that the woman in the following video would be one of the most cognitively successful problem solvers in the real world. My guess is that she spends much of her time googling stuff to help her solve problems for much of the day.

    Great post, thought provoking. Thanks

    • Doug Lemov
      November 13, 2017 at 7:08 pm

      we read the cognitive science to say very different things you and I.

Leave a Reply