In the first of a four-part series, Mark Enser recalls feeling dubious about a technique that looked suspiciously like rote learning. But after analysing the evidence, he became an enthusiastic convert
The term “retrieval practice” burst on to the mainstream education scene a few years ago. Like many ideas that suddenly arrive fully formed and without explanation, it put my natural cynicism to work. After all, we’ve been stung before: Brain Gym, learning styles, minimal instruction and the “guide on the side”, to name just a few. Could retrieval practice be the next fad to come, create work and go?
On the face of it, the idea seemed too good to be true: give pupils regular quizzes so that they retrieve information studied in the past and they will learn more. It couldn’t be that easy. Perhaps retrieval practice would just help them to learn key facts by rote but lead to nothing more meaningful.
Being able to recall the death toll of the Nepal earthquake or the pressure system on the equator is great, but of limited use if you can’t do anything with that information.
My desire to see if retrieval practice could lead to meaningful learning led me to discover Karpicke and Grimaldi’s paper “Retrieval-based learning: a perspective for enhancing meaningful learning” (2012).
This overview of the research into retrieval practice has helped to shape my attitude towards not only the humble quiz but also the nature of learning. It has helped me to teach my pupils how to revise more effectively and it is the one paper I would recommend that every teacher reads.
Changing ideas on learning
Karpicke and Grimaldi start by considering the question of what we mean when we talk about “learning”. They introduce one of the simplest and most useful definitions I have come across, saying that “if a person has learned something, it means they are capable of using information available in a particular context, referred to as ‘retrieval cues’, to reconstruct knowledge in order to meet the demands of the present activity”.
They go on to explain that, for too long, learning has been considered something that happens during studying, when the material is “encoded”. This, they suggest, is why we often use metaphors for learning that involve physical buildings. We construct knowledge through learning structures and seek to understand the architecture of the mind.
Critically, they make the point that we need to turn this on its head and realise that meaningful learning doesn’t happen just through the studying of new information but through its retrieval over time.
Testing is not neutral
Retrieval is important because every time we bring something back into our working memory, we change the original memory. We don’t access perfect copies of the original but rather versions shaped by what we have learned since. By retrieving material, we reconsider it in the light of new information.
This enables us to build complex webs of learning that we call “schema”. For example, when I retrieve the death toll of the Nepal earthquake, I might be able to connect it with what I have since learned about problems with building codes in the capital city and the links to increased urbanisation and rising populations.
When I recall that the equator is dominated by low pressure, I might be able to connect this to the convection rainfall and tropical ecosystem found there.
This is why Karpicke and Grimaldi argue that testing is not, as was once believed, a “neutral act” – a way of finding out what was learned. Instead, they say, it is an intrinsic part of the learning process.
In their paper, “Ten benefits of testing and their application to educational practice” (2011), Roediger, Putnam and Sumeracki discuss the positive washback that is achieved from testing or quizzing pupils beyond the assessment itself. These benefits include the organisation of material and the increased durability of the memory (its storage and retrieval strength).
These ideas on increased durability of memory have also shaped how I advise my pupils to revise. In the past, I encouraged them to do this through re-studying material: reading their book and making notes; reading a revision guide and answering questions; making a mindmap based on their notes.
What has changed my approach is a study cited in Roediger and Karpicke’s 2006 paper “Test-enhanced learning: taking memory tests improves long-term retention”. The authors conducted an experiment in which three groups of students were presented with information on the same topic.
The first group studied the information on four separate occasions, the second group studied it the first three times but in the last session tried to recall as much of it as they could from memory. The third group studied it only on the first occasion, and in the three following sessions tried to recall as much as they could. They never saw the original material again.
Perhaps unsurprisingly, the first group were the most confident before the final test. The third group, who had seen the material only once, were the least confident going in. The results were far more unexpected. The group who had seen the information only once, and then tried retrieval practice, performed significantly better on the test than the groups who had studied it more often. The group who had only studied it and had not tried to recall it did the worst.
I now suggest pupils review their notes, then put them away and try any tasks from memory before checking back afterwards to see how they did. In my experience, they will not do this out of choice (we assume we will do better through re-study than retrieval and have to be convinced otherwise).
This paper has shaped my practice as a teacher. It has put my concerns about rote learning to rest and helped me to see how testing can lead to meaningful learning, where pupils can apply what they have learned to different situations.
It’s also had an impact on my pupils, who now have a much better idea of how to maximise their time spent revising. They do less of it but do it better, and we are all much happier as a result.
Mark Enser is head of geography and research lead at Heathfield Community College. His first book, Making Every Geography Lesson Count, is out now. He tweets @EnserMark
This is the first in a four-part series on the research that has influenced teachers’ careers. Next week: scrutinising observations
This article originally appeared in the 9 August 2019 issue under the headline “Retrieval practice: far more than a fad”