How can ‘mixing it up’ and confusing students be good for them?

When we plan an activity, we instinctively group like with like. We discuss Monet, then talk about Renoir; we add fractions, then move on to subtracting them. Sometimes, however, mixing things up can help. 

It’s called interleaving, and it’s a desirable difficulty: a way to make learning harder initially, that boosts student retention subsequently. Some desirable difficulties are pretty straightforward. Spacing learning, for example, just means… well, spacing learning out by revisiting topics.  

Interleaving is a bit more complicated though, both because it’s counter-intuitive, and because researchers note that we sometimes apply it in ways that go beyond the evidence. 

A classic study illustrates interleaving well. Researchers taught students how to find the volume of four 3D forms (obscure ones, to reduce the influence of prior knowledge). Students then did 16 practice questions, four for each form. Some students – blockers – practised like with like (four spherical cone problems, then four spheroids, and so on). Others – mixers – practised the same questions ordered at random.  

Initially, blockers did better than mixers because the mix was confusing. But when they returned for a test a week later, mixers did three times better than blockers. As the researchers noted, simply changing the order of practice “led to large gains in test performance”. 

So why does interleaving work? The researchers suggested that mixing things up “requires that students learn not only how to perform each procedure but also which procedure is appropriate for each kind of problem”.  

It’s easy to imagine students asked to add fractions completing 20 questions without focusing fully on each one; they just apply the same method each time. On the other hand, if each question type differs, students have to think harder. They are likely to make more mistakes early on, but when they encounter a new question in the wild, they are more likely to be able to pick a good approach. 

In a helpful review of the science of learning strategies, Weinstein, Madan and Sumeracki describe three situations in which interleaving can help students think harder, and so perform better:  

• Problem solving, such as mixing addition and subtraction problems, or drilling a variety of physical skills rather than a single one. 
• Learning to distinguish between categories, such as attributing paintings to artists. 
• Alternating between models and questions, rather than just showing a model, then asking students to complete questions. 

However, there are two caveats about interleaving.

First, one elegant study demonstrated that students aren’t keen on it and don’t think to do it unless prompted. Students were learning to associate artists with their artworks. Different groups of students encountered artists and artworks in five different ways, from purely interleaved to purely blocked. Mixing up the order of the artists meant students learned them better. But asked whether they preferred interleaving or blocked learning, students tended to prefer blocking. 

The article suggested neat ways to compromise, like mixing up blocks, or becoming increasingly random. These had the same positive effects as interleaving, and students were keener on them. But it’s worth noting the researchers’ finding that students rarely think to interleave, unless we suggest it to them. 

The second caveat is that it doesn’t work for everything. For a start, it has mostly been tested in maths and science. There’s nowhere near as much evidence for its effectiveness in other subjects.

And then, as Weinstein and her colleagues note, interleaving has only been tested for learning within lessons. So while some suggest interleaving across the curriculum or between subjects, this isn’t something for which we have evidence. A little confusion helps, but too much doesn’t – particularly if it discourages students. 

Interleaving improves retention, but there are other great ways to improve retention: what’s best depends on what is being taught and to whom. So, as with many intriguing research findings, once we’ve discussed the evidence, we return to the most important factor for successful implementation. Namely, that there is almost nothing that can be prescribed for all classrooms, and the best person to decide whether, how and when to apply a strategy is the evidence-informed teacher.