There's no doubt that teachers are becoming more interested in the evidence that underpins learning.
This sometimes means, though, that they discover that what they thought was effective is, in fact, not. The past few years have seen much conventional wisdom questioned and strategies cast aside, leading to a fundamental shift in the way teachers teach.
Distributed practice is one such disruptive theory.
Generally, topics are still taught in blocks as a massed practice form of teaching, covering the content and then assessing knowledge and understanding on completion before moving on to the next topic.
The problem with massed practice is that the information learned is often not accessed for long periods, making forgetting it more likely.
Research seems to suggest that spacing learning out over a longer period of time is more efficient and improves outcomes in terms of being able to recall the information taught – so-called distributed practice.
Quick read: Eight key pieces of research for teachers
Quick listen: Why we need more hands-on learning in schools
From the magazine: How ‘journal clubs’ can help teachers to evaluate pedagogical research
Distributed practice, therefore, constitutes desirable difficulty, a task that requires substantial but necessary effort to complete. You can see how this works in reality: rereading notes, for example, doesn’t constitute a desirable difficulty because more effort is going into reading the words than processing the information. Flashcards, however, require a student to actively recall the information.
Distributed practice: how it (seems to) work
Why distributed practice leads to better retention is not fully understood and there are several possible explanations. The most plausible explanation is that a time delay before restudy changes the context, boosts attention and allows for time to consolidate the information through sleep.
It also allows time for forgetting, which may ultimately (and perhaps counterintuitively) lead to better retention.
The choice of the time gap between study sessions is also somewhat debatable, but it does appear that the longer the interval, the more positive the outcome.
Despite some criticisms, however, the research evidence supporting distributed practice is pretty robust. Nevertheless, studies have tended to take place in laboratory settings where the environment is tightly controlled, unlike the classroom where distractions are common.
Does it really work?
For this reason, many of these studies lack ecological validity – ie, they don’t represent what happens in the real world. Also, these studies most often focus on declarative memory, such as memorising lists of words or foreign language learning.
A new study attempts to tackle some of these weaknesses by examining procedural memory in a real-world setting. Procedural mathematics learning differs from memorisation in that it involves the ability to execute action sequences (or procedures) to solve problems.
Katharina Barzagar Nazari and Mirjam Ebersbach, of the University of Kassel in Germany, recruited 213 third- and seventh-graders and taught a maths lesson that was developed from their regular curricula. Students were required to learn facts taught in the classroom and also worked on slightly different but conceptually similar maths exercises that were massed or distributed. The massed practice group worked on three sets of exercises in one day while the second group worked on one set for three days. They then completed follow-up tests at one and six weeks.
The expectation was that students in the distributed practice group would outperform the massed practice group. It was also hypothesised that the differences would be even more marked six weeks later due to previous studies finding that the positive effect of distributed practice increases over time.
Results were certainly encouraging but didn't quite match expectations. Overall, distributed practice did lead to better retention in the short term. However, there was no significant improvement at the six-week time point, even though grade 7 students showed greater stability than grade 3. These latter findings, therefore, also run counter to previous research.
A particular weakness of the study is the low sample size of just 213 participants. However, the results were examined using a technique known as Bayesian analysis, which is thought to be more appropriate for use with smaller samples. Nevertheless, it will be interesting to see if these results remain consistent if the study is replicated with larger samples.
Of course, the big question for teachers is whether this particular study implies that they should replace massed practice with distributed learning. Barzagar Nazari and Ebersbach do conclude by indicating that distributed practice provides a promising strategy in the context of classroom learning, but does this mean that teachers should start using it in their classrooms?
In a 2014 paper, Professor Carolina Küpper-Tetzel stated that “future studies will have to fine-tune theories to strengthen the significance of empirical results and to allow for better recommendations to educators".
Perhaps the simple answer is that introducing distributed practice is unlikely to have any negative impact on learning and could also provide a useful and worthwhile project for teachers interested in carrying out their own small-scale research studies.
- Barzagar Nazari, K, and Ebersbach, M (2019) "Distributing mathematical practice of third and seventh graders: Applicability of the spacing effect in the classroom", Applied Cognitive Psychology, 33(2), 288-298
- Küpper-Tetzel, CE (2014) "Understanding the distributed practice effect: Strong effects on weak theoretical grounds", Zeitschrift Fur Psychologie/Journal of Psychology, 222(2), 71-81
Marc Smith is a chartered psychologist and teacher. He is the author of The Emotional Learner, and co-author with Jonathan Firth of Psychology in the Classroom. He tweets @marcxsmith