Studies of retrieval practice have revealed that retrieving some information from memory can impair subsequent recall of other information, a phenomenon often termed retrieval-induced forgetting.
Research on the detrimental consequences of retrieving some items from memory has a long tradition which has produced a base of literature that is now replete with rich empirical and theoretical reviews that explore how and why retrieval-induced forgetting has been observed in a wide variety of contexts (see, for example, Storm et al, 2015).
Knowledge of how retrieval practice may have both enhanced and inhibited recall of information from memory is clearly useful to teachers. Where relevant it makes sense for them to draw on the findings from this research to help them in their daily work.
An often-overlooked aspect of research
In general, the most helpful research evidence is that which is sufficiently nuanced, comprehensive and transparent to capture both the potentially beneficial and potentially detrimental effects of a particular practice or process.
Curiously though, information relating to retrieval-induced forgetting is noticeably lacking in key education policy and guidance documentation, not least Ofsted’s overview of research evidence underpinning its Education Inspection Framework (Ofsted, 2019), the Department for Education’s Early Career Framework and ITT Core Content Framework (DfE, 2019), and the Education Endowment Foundation’s Cognitive Science in the Classroom review (Perry et al, 2021).
All of these documents frequently advocate the use of retrieval practice, yet none of them discuss appropriately, and generally not at all, details of how researchers have concluded that retrieval can impair the recall of information.
And while they include references to research relating to the expressed benefits of retrieval, they do not direct readers to research sources concerned with retrieval-induced forgetting.
The omission of sufficient discussion of the recorded adverse effects of retrieval practice is not restricted to education policy and guidance, it is prominent in several high-profile reviews of retrieval practice.
Reviews by Roediger and Karpicke (2006), Rowland (2014), and Adesope et al (2017) dedicate some space to consideration of the reported effects of retrieval on forgetting and include references to a range of supporting literature, nonetheless the implications of this research do not feature in, and thereby do not influence, the conclusions of their reports.
Consequently, it is possible that, until now, many readers will not have read any research material relating to retrieval-induced forgetting and may not even have heard of it.
So, what is retrieval-induced forgetting?
In education settings, retrieval practice research has been primarily, but not entirely, concerned with investigating the effect of practising the retrieval of some information from memory on the recall of that information at a later time.
The basic underlying approach of this research has been to ask participants in an experiment to study a categorised selection of “to be remembered” material and then repeatedly practice retrieving half of the categories studied.
After a delay, participants take a test covering all the categories studied. If participants recalled more of the practised than unpractised material, then researchers have drawn the conclusion that practising retrieval enhanced the subsequent recall of the practised items.
Research on retrieval-induced forgetting has focused on the effect that practising retrieval has had on some of the unpractised material. Typically, in this type of research only half of the items in each of the categories selected for retrieval are repeatedly practised. This procedure creates the additional opportunity to examine recall of both practised items and unpractised items from a category.
Put concisely, this research has revealed that students recalled unpractised items from practised categories less well than other unpractised items at a subsequent test. This finding is referred to as retrieval-induced forgetting
An illustrative example of retrieval-induced forgetting
The results from influential research by Anderson et al (1994) depict this finding quite clearly. The study reported the following recall success rates from an experiment using the standard retrieval-induced forgetting paradigm:
- Recall of practised items from studied categories: 73.6%
- Recall of unpractised items from practised categories: 37.5%
- Recall of other unpractised items: 48.4%
In this particular instance, the recall success rate for practised items was 25.2 percentage points (73.6% vs 48.4%) higher than the general unpractised rate. This result offers support for the view that practising retrieval can enhance the subsequent recall of the practised items.
It is crucial to note, however, that the recall success rate for unpractised items from practised categories was 10.9 percentage points (37.5% vs 48.4%) lower than the general unpractised rate. This result offers support for the view that recall of unpractised items from practised categories can be impaired at a subsequent test.
Considering these two results together shows how the overall potential benefit of practising retrieval became attenuated, from a 25.2 percentage points gain to a 14.3 percentage points gain (25.2% – 10.9%). In everyday terms it might be said that on this occasion retrieval-induced forgetting reduced the impact of retrieval practice by more than 40% (10.9/25.2 x 100%).
What does this mean for using retrieval practice in your teaching?
If teachers and leaders are to make the best evidence-informed choices about whether to select retrieval practice over alternative teaching approaches for use in their classrooms and schools, then it is important for them to gain, at the very least, a broad awareness of both the reported beneficial and detrimental aspects of practising recall from memory.
This knowledge can then be used to assist teachers and leaders in devising and deploying retrieval activities that have tangible potential to both enhance recall from memory and at the same time minimise forgetting.
For sure, teachers can take advantage of the stated benefits of practising retrieval to develop their classroom activities with the goal of helping pupils and students recall some information when it is required.
Yet, it is important to realise that the effectiveness of those activities may be overshadowed somewhat by an increase in the forgetting of other important information which is brought about by the retrieval process.
As research on retrieval practice progresses the outcomes from it will become more informative for developing our understanding of retrieval practice and retrieval-induced forgetting.
It is worth noting though that so far, research focused on retrieval-induced forgetting in education settings has highlighted the need for care and caution to be taken when developing retrieval practice activities. Presently, it seems evident that practising retrieval has potential to improve student performance in subsequent tests that require recall of the same information as that practised.
It seems equally clear that if the items on a subsequent test are unlikely to be identical, or at the least very similar, to those which have been practised it is unlikely that the potential benefits of practising retrieval will be realised fully.
Moreover, and perhaps most importantly, if the test is likely to contain unpractised material which is semantically related to that which was practised then recall performance on this material is likely to be impaired.
Comments on retrieval practice research
So, when considering the existing evidence-base on retrieval practice, teachers and leaders may profit from giving some consideration to the following aspects of the research:
First, solitary effectiveness studies of retrieval practice have recorded mixed results. While positive results have been reported, null effects, where practising retrieval has had little effect on the recall of practised items, and reverse effects (sometimes referred to as the “negative testing effect”) where practising retrieval has led to impaired recall of the practised material, have also been recorded. Case studies are therefore most appropriate for developing an understanding of what happened when a distinct intervention was applied at a particular time in a specific context.
Second, reviews of retrieval practice studies are not always sufficiently comprehensive or representative. As noted previously, it is not uncommon for writers to have predominantly cited studies that reported positive effects of retrieval and to focus their discussions mainly, and at times entirely, on the declared benefits of practising recall from memory.
Third, a large proportion of retrieval practice research has used a very specific experimental paradigm which utilises cued recall of a single word. These studies have used prompts to support retrieval during the practice phase of the experiment and to support recall at the final test stage. Cued recall of the type used in this research may not be the foremost method of retrieval practice that is being planned for use in an individual school or classroom. Therefore, teachers and leaders are likely to benefit most from consulting research that has been undertaken using approaches along the lines of that they intend to deploy themselves.
Fourth, unsurprisingly, the mixed results from empirical studies have prompted researchers to explore how retrieval practice may have both benefited and hindered recall from memory. These theories have been, and continue to be, refined, extended and challenged as researchers test them against numerous empirical elements of retrieval-induced remembering and forgetting. However, as with many education phenomena, it is unlikely that any single model can account comprehensively for the breadth of empirical findings reported in retrieval-induced forgetting research.
Conclusions
In closing, it is worth noting that nuanced and mixed results, like those recorded in retrieval practice research, are pretty commonplace in investigations of education phenomena. At times this can be confusing to readers and leave them with a sense of not knowing which results to take seriously.
Moreover, eye-catching proclamations of the effectiveness or ineffectiveness of education interventions can be very enticing to readers and pounced on as evidence to support a particular call to action.
However, high-quality research in education settings is generally better at describing what happened than predicting what will happen. Although the value of each study needs to be assessed on its own merits, undeniably it is not possible to know in advance which practices will work best at a given time in a school or a classroom.
This situation obviously doesn’t prevent teachers and leaders from deciding what to do, but it does highlight one of the limitations of evidence-based practice. Consulting the literature can be very helpful but it is unlikely to provide a comprehensive solution to a practice quandary.
Teachers and leaders must therefore be able to use their professional judgement about what to do in their classrooms and schools, and findings from research can help formulate those judgements sensibly.
Reflective questions for teachers/leaders
- How highly would I rate my current understanding of retrieval-induced forgetting?
- Am I drawing on material from a suitably comprehensive base of research about retrieval-induced forgetting to develop my understanding?
- In what ways will a good understanding of retrieval-induced forgetting help me to design retrieval practice activities?
- Have I got sufficient information to suggest that retrieval practice is likely to be one of the most appropriate activities to use at this particular point in my curriculum/lesson?
- What has been reported in retrieval-induced forgetting research about the type of retrieval practice I am planning to use?
- What consideration have I given to the potential of creating retrieval-induced forgetting when using this type of retrieval practice at this particular time?
- What steps am I planning to take in order to minimise the effects of retrieval-induced forgetting when using retrieval practice this way?
- Terry Pearson is a former teacher, senior leader, teacher educator and researcher having a career in education that has spanned more than 35 years. He has always possessed a keen interest in education research and now spends much of his working time as a meta-researcher.
Further information & resources
- Adesope et al: Rethinking the use of tests: A meta-analysis of practice testing, Review of Educational Research (87,3), 2017.
- Anderson, Bjork, & Bjork: Remembering can cause forgetting: Retrieval-dynamics in long-term memory, Journal of Experimental Psychology: Learning, Memory, and Cognition (20), 1994.
- DfE: Early Career Framework, January 2019: https://bit.ly/3sXPGKh
- DfE: ITT Core Content Framework, 2019: https://bit.ly/3HvuRi8
- Ofsted: Education Inspection Framework: Overview of research, January 2019: http://bit.ly/2Vgzt2z
- Perry et al: Cognitive science approaches in the classroom, Education Endowment Foundation, July 2021: https://bit.ly/3FjLueC
- Roediger & Karpicke: The power of testing memory: Basic research and implications for educational practice. Perspectives on psychological science (1.3), 2006.
- Rowland: The effect of testing versus restudy on retention: A meta-analytic review of the testing effect. Psychological bulletin (140,6), 2014.
- Storm et al: A review of retrieval-induced forgetting in the contexts of learning, eyewitness memory, social cognition, autobiographical memory, and creative cognition, Psychology of Learning and Motivation (62), 2015.