I’m aware that starting an article with a film reference is a bit of a cliché, but please bear with me.

In the film The Matrix, Morpheus tells eventual-saviour-of-the-world Neo: “Sooner or later, you’re going to realise, just as I did, that there’s a difference between knowing the path and walking the path.”

It’s a quote that will resonate with many teachers. As we fight our individual battles to make the world a better place, we see only too well that there is a gap between knowing, in theory, what is the right thing to do and how we make that work in practice.

Increasingly, “knowing the path” in teaching seems to involve being “evidence informed”. In other words: teachers should be aware of how to translate the results of research into successful learning and be utilising these insights in their own practice. 

Quite right, too, you might say, and I would agree. However, becoming evidence informed is not without its challenges and requires well-meaning teachers to walk a tightrope between two suboptimal outcomes.

On the one hand, if research is interpreted too simply, crude interventions based on poorly understood mechanisms can cause more problems than they solve. Such instances are often termed “lethal mutations” and seem especially likely to occur when schools or trusts attempt to condense extensive research findings into a small number of codified “best practice” prescriptions.

Under such circumstances, effective ideas and good intentions can become twisted into practices that are, at best, ineffective and, at worst, harmful to student progress.

Becoming evidence informed requires well-meaning teachers to walk a tightrope between two suboptimal outcomes

On the other hand, entirely faithful applications of research may not be practically possible, or desirable, in a complex school environment. For example, much of the cognitive psychology research into learning has been conducted in environments that are highly controlled, such as dark or bare laboratory rooms, using “controlled” stimuli, such as nonsense syllables, which bear little resemblance to the context of learning in school environments. 

Even studies that have been more “ecologically valid” - as psychologists term settings or tasks that more closely resemble how the skills being studied occur in real life - will often use samples of participants (such as college students) who are very different from the pupils in our everyday classrooms.

But suppose we find research that jumps all these hurdles? Then we have a further problem. 

Research studies have only a singular aim, or a very limited number of aims, allowing them to isolate the specific effects of one variable. Teachers do not have this luxury. Our classrooms are uncontrolled and noisy (in the statistical sense of being highly variable, though sometimes just plain noisy, too). We have to consider a huge range of interconnecting variables concurrently every time we plan any activity, and this reduces our ability to quantify the effect of any single change. True fidelity to the research is, therefore, likely to be an unattainable ideal.

Where, then, is the sweet spot for a teacher attempting to apply the principles of research without, on the one hand, propagating crude lethal mutations or, on the other hand, becoming overly entangled in the finer details?

I put this question to a number of research experts in different areas of the science of learning (memory, attention and metacognition) to examine how, or if, we can walk the evidence tightrope between simplicity and fidelity.

Memory: retrieval practice

Retrieval practice (getting pupils to bring information to mind from memory) is an increasingly common feature of many teachers’ classroom practice - not least because the approach features in official frameworks, such as Ofsted’s research for the education inspection framework and the Department for Education’s core content framework for initial teacher training.

The strength of the research findings and the promise of improvements in students’ retention of information (a recent Education Endowment Foundation review of cognitive science approaches in the classroom found that the evidence for retrieval practice quizzing “is moderate and generally positive”) have led many schools to incorporate retrieval into their teaching. 

However, even with such a seemingly robust effect, there is still an evidence tightrope to walk. An oversimplified message that “quizzing is good” may lead teachers to ignore the retrieval element, resulting in activities that do not actually require students to retrieve information from memory (for example, looking up answers they are unsure of). 

What’s more, a recent review of factors that can affect retrieval practice in the classroom, by Michael Trumbo and colleagues (2021), found that retrieval may only benefit learners once they have an organised, integrated mental representation of information. 

Quizzing learners who still don’t understand or have no clear memory of the material to start with is therefore unlikely to be helpful.

But if “just give them a test” is too simplistic, how much more detail do teachers need to know? 

While Trumbo and colleagues conclude that the benefits of testing are “robust and generalisable” enough to make the practice “worthy of inclusion to at least some degree in most educational contexts”, they also lay out the many other variables that have also been found to moderate this positive effect. 

These include the format of the test (for example, multiple choice, short-answer questions or longer essays), the amount and type of feedback given, the retention interval, the number of re-study and quiz attempts permitted, the complexity of the information, student ability, prior knowledge and motivation, metacognition and metamemory (students’ own judgements about their learning and the strength of their memories).

It’s a lot to take in. While understanding too little can be problematic, if teachers read too much into all of these different variables, a sort of evidence paralysis can set in, leaving them dithering over strategies in much the same way that I do over all the different kinds of jam in the supermarket. 

So, where is the conscientious educator to start? Professor Mark McDaniel, who co-authored the paper with Trumbo, and also co-authored the book Make it Stick, believes that “teachers can get started with implementing testing for learning without worrying too much about the nuances”, while still acknowledging that those nuances exist.

In a forthcoming paper, co-authored with colleagues (Pan et al, 2022), he provides four recommendations, which serve as a useful starting point for using retrieval practice. 

• Students should engage in genuine, effortful attempts to retrieve information from memory. 
• Retrieval should typically be low stakes. 
• Retrieval should include feedback that contains the correct answers. 
• It should be repeated on a number of occasions. 

Given these starting points, however, McDaniel is keen to stress the role of professional judgement in each teacher’s development of retrieval routines in their classrooms. 

“It could well be that teachers have to try an implementation that fits their logistical, practical and academic demands, and then fine-tune if needed,” he argues. 

Attention: perceptual load

Of course, retrieval of information isn’t going to happen if it is never committed to memory in the first place. 

Nilli Lavie, professor of psychology and brain sciences at University College London (UCL) and a world expert in the study of attention, is keen to stress just how limited our attention system is. 

As she explains: “We have a very limited capacity for the perception of information. If you don’t perceive, you don’t process further. That is absolutely vital to understand, so that we can create the conditions where the information that is conveyed by the teacher is not superseded.”

However, understanding that we have capacity limits can also invite lethal mutations: Lavie is concerned that learning this information in isolation might lead teachers to focus too heavily on trying to reduce perceptual load (providing information in such a way that it does not burden attention). 

Conversely, “one of the crucial things for focus”, she says, “is to fill capacity as much as possible”. If we fill our capacity with task-relevant information - in other words, if the main focus is made as attention-grabbing as possible - this reduces our ability to process other (often task-irrelevant) information from the environment. 

Aiming to fill students’ perceptual capacity may therefore be a good starting point for increasing their focus, but, Lavie warns, this needs to be through information directly relevant to the task or material being learned. 

Funny cartoon GIFs dancing at the side of your PowerPoint slides will be unlikely to illustrate the key idea

Take visual cues, for example. We know that these can be “a good way to engage attention”, Lavie says. But, she adds, any visual aids used to capture students’ attention “should be a good illustration of the principle that is being taught”. 

It’s not just about filling students’ perceptual capacity, then; we also need to think about what information we are filling it with. Funny cartoon GIFs dancing at the side of your PowerPoint slides will be unlikely to illustrate the key idea. Instead, they are likely to be a distraction.

Teachers having no knowledge of attention research can be problematic. But, says Lavie, knowing a lot about one area of that research without sufficient understanding of the other areas could be just as damaging as knowing nothing at all.

She gives the example of gaze orientation. This can be a controversial topic in education, as “track the speaker” - a system in which students follow the speaker in the classroom with their (usually visual) attention - is a staple part of hotly debated classroom behaviour programmes such as SLANT (sit up, lean forward, ask and answer questions, nod your head and track the speaker) and STAR (sit up, track the speaker, ask and answer questions, respect those around you). 

Lavie describes two lines of research, one of which could, in isolation, be taken to support “track the speaker” methods and another, which could be used to argue the opposite. 

In the first, she explains, we know that “when people orient attention, they also move their eyes to the direction of their attention”. 

“Multiple sources of the evidence show this strong link. This suggests that monitoring visual attention is likely to be a simple way for a teacher to gauge attention and to increase the chances of student focus,” she says. 

“[However], a second line of research shows that while, typically, attention and gaze follow each other (and vice versa), you still can orient attention away from that. Especially when we are under high working-memory load (when thinking hard), we may orient our gaze away to better concentrate on the internal processing of the information.”

Ultimately, she adds, in this case, “over-interpreting one line of research can lead to people demanding that all pupils orient their gaze” when, in fact, the full picture is not nearly so simple. 

Metacognition

Having an awareness of how different lines of research interact isn’t only important within fields; in some cases, it’s also necessary for teachers to understand how different fields speak to one another.

Take metacognition, for example. Helping students to develop their metacognition (the awareness and understanding of our own thought processes and the patterns behind them) is one of the most impactful strategies listed in the EEF’s Teaching and Learning Toolkit. When done well, the toolkit states, it can be worth the equivalent of an additional seven months’ progress. 

“As with all things, though, seeing metacognition as a magic bullet, without bothering to engage with the evidence, is going to greatly reduce the effectiveness of what you are doing,” warns Jonathan Firth, a teaching fellow at the University of Strathclyde’s school of education.

“If educators want to make their learners more metacognitively sophisticated, then they themselves also need to understand the learning processes that are happening,” he continues. “If you want learners to reflect on their learning and plan independent study, then they need to be taught something about how learning works, about how forgetting works and so on.” 

In other words, if teachers have an overly simplistic model of memory and attention, this will likely impair any attempts they make to develop their students’ metacognition. 

By extension, excessive complexity or a blinkered dive into one particular line of research is also likely to impact the ability of teachers and students to think about their own decision making and learning in a useful way. 

However, Firth thinks that the risk of teachers attempting to follow research too faithfully or not faithfully enough is less pronounced here than in some other areas of pedagogy.

“There’s not the same danger [compared with memory and attention] of people getting things really wrong and doing something counterproductive if they don’t understand the evidence,” he says. “The biggest risk is that [work around] metacognition will simply not be very effective if not done well.”

Knowing a lot about one area of research without sufficient understanding of the other areas could be just as damaging as knowing nothing at all

So, what does all this mean for how teachers approach education research? Are there any ground rules that will help them to translate it into practice?

Although the exact starting point may vary between disciplines, the experts I spoke to had some general advice on how to keep your footing while walking that evidence-based tightrope.

1. Keep it simple, but know your ‘why’

The first point of advice is to start simple and know why you are doing something. A useful tip is to check if you can explain, out loud, why something is likely to work. If you can, it means that basic applications are likely to be on the right lines. 

When it comes to retrieval practice, for example, Professor Henry Roediger, McDaniel’s Make it Stick co-author, suggests that the simplest applications are often the most useful.

“At the beginning of class, teachers could ask [retrieval] questions about the previous class rather than (or in addition to) reviewing the content. And every 15 minutes or so during class, the teacher could ask students ‘What important things have you learned so far today?’ That practice would only take a few minutes of class time and would help to consolidate the day’s learning,” he says.

2. Reflect and adjust

Rather than agonise over the minutiae of different research effects, your time may be better spent reflecting on what went well for you in your own setting and how you could improve. 

David Shanks, professor of experimental psychology at UCL, lays out how a process of iterative improvement from a baseline of an initial “research-based” strategy might look. 

He suggests that “research can be built into classroom practice in a way that preserves what are thought to be its key ingredients, as much as possible”. 

“This intervention can then be piloted,” he continues. “The piloting will reveal what the unexpected issues created by the implementation are as well as providing first-hand experience of whether the intervention is effective (and perceived to be so). If there are problems, then the intervention needs to be adapted and re-piloted - though, of course, doing so may dilute its effectiveness.” 

3. Trust yourself

Teachers have enormous amounts of tacit expertise about the conditions under which effective learning happens, and we can sometimes, as a profession, lack the confidence to factor this into the equation. 

Lavie points out that “if something that you learn from science does not make sense, given what you see in the classroom”, that particular approach might be better left to the laboratory, at least for now.

“There are a lot of lines of research and you’ll never know all of what exists in science. Even the researchers don’t know all of it,” she says.

None of this is to say that we don’t need to move forward on the finer details of classroom-based research. There remain many important applied questions that hold great promise for education. It’s just that, according to the experts, we can all have some confidence in doing the basics right while we wait for the answers to the more complex questions to be worked out. 

Walking the path between oversimplification and the sea of complexity may be easier than we might have imagined, then. All we need to do is start simple (but know why), reflect and adjust, and trust ourselves. It’s advice that even saviours of the world like Neo could benefit from.

Michael Hobbiss is a psychology teacher and cognitive neuroscience researcher