This article was written for SecEd magazine and first published in June 2018. You can read the original version on the SecEd website here.
You can access the full archive of my columns for SecEd here.
Research by Kirschner, Sweller and Clark (2006) compared guided models of teaching, such as direct instruction, with discovery learning methods, such as problem-based learning, inquiry learning, experiential learning, and constructivist learning, and found that the latter methods didn’t work as well as the former.
It didn’t matter, they argued, if pupils preferred less guided methods, they still learned less from them (see also Clark, 1989).
In his book, Visible Learning, Professor John Hattie found that the average effect size for teaching strategies which involved the teacher as a “facilitator” was 0.17, whereas the average effect size for strategies where the teacher acted as an “activator” was 0.60.
Direct instruction had an effect size of 0.59 compared to problem-based learning with an effect size of just 0.15.
Therefore, direct instruction – it seems – is more effective than discovery learning approaches. But what, exactly, does good direct instruction look like in practice?
Personally, I think direct instruction works best when it follows this four-step sequence:
Telling – or teacher explanation – works best when the teacher presents new material to pupils in small “chunks” and provides scaffolds and targeted support.
Showing – or teacher modelling – works best when the teacher models a new procedure by, among other strategies, thinking aloud, guiding pupils’ initial practice and providing pupils with cues.
Doing – or co-construction – works best when the teacher provides pupils with “fix-up” strategies – corrections and “live” feedback.
Practising – or independent work – works best when the teacher provides planned opportunities in class for extensive independent practice.
Of course, the learning process does not end here. Rather, pupils need to garner feedback on their independent practice and then act on that feedback in order to improve by increments. We’ll return to the importance of this “feedback loop” later. But first, let’s take a look at each of the four steps in our teaching sequence.
The most effective, expedient way for pupils to acquire new information is for the teacher – that educated, experienced expert at the front of class – to tell them what they need to know.
This is not to suggest that sometimes, for some purposes, other approaches are not also effective, but teacher explanations remain the most efficient method of teaching – not to mention the least likely to lead to misconceptions among pupils and a misunderstanding by the teacher of what pupils can and cannot do. So what are good explanations made of?
First, good explanations involve metaphors and analogies because this enables the teacher to contextualise new information so that abstract ideas or hitherto alien concepts are made concrete, tangible, and real, and so that they are related to pupils’ own lives and experiences.
Second, good explanations make effective use of dual coding. In other words, teachers’ verbal instructions, as well as any text-based explanations displayed on the board or in handouts are paired with and complemented by visuals such as diagrams, charts, graphics and moving images.
And finally, good explanations are reciprocated, with pupils explaining concepts back to the teacher as well as to each other. This works on the basis that only once you teach something have you truly learned it. Learning by teaching works because, by teaching, pupils gain feedback and make better sense of a topic. Learning by teaching also works because it is a form of learning by doing, of practising, and thus provides a source of both intrinsic and extrinsic motivation.
Once teachers have explained something, they should make effective and plentiful use of models – exemplars of both good and bad work, as well as exemplars from a range of different contexts – which show pupils what a final product should look like and what makes such products work.
Good models demonstrate what works as well as what doesn’t. It is important to show pupils what excellence looks like by sharing models of the very best work, giving them something to aspire to, and an understanding of how to produce high-quality work of their own.
But it is equally important to show pupils models of ineffective work, work that isn’t quite the best (or perhaps is so very far from being the best) so that pupils can learn what not to do and how to avoid making the same mistakes themselves.
All the models that are shared should be dissected in front of pupils, with the teacher demonstrating the dissection process.
For example, if a model of a persuasive speech is shown on the board, the teacher should analyse it using text marking, pointing out and then annotating how it works, what makes it effective, breaking it apart to identify and discuss each of its component parts. Then the teacher should reconstruct the speech, explaining how the component parts hang together to create an effective argument, how the whole becomes something much greater than the sum of its parts.
Once pupils know how to dissect models, they should be afforded the opportunity to do so without the teacher’s guidance, perhaps by teaching other pupils. In order to prepare pupils for this, it is important that the teacher offers encouragement, gives specific instructions, uses thought or sentence stems to provide pupils with the right language, and – as I say above – directly demonstrates the process first.
Once the teacher has modelled something at the front of class, it is important to do so again but, this time, with pupils’ help. Co-construction (or joint-construction) works well because the teacher engages pupils’ thought processes and helps them by questioning their decisions and by prompting further decision-making.
The teacher’s role is not to construct another model herself but to ask targeted questions of pupils to encourage them to complete the model together, as well as to provide corrections and feedback along the way, and drip-feed key vocabulary into the mix.
For instance, and to return to the example above, if a teacher has explained to a class how to write a persuasive speech and then modelled doing so on the whiteboard while thinking aloud, she might then ask the class to produce a persuasive speech of their own.
The teacher may begin by asking the class to determine an audience and purpose for the speech, then prompt pupils to debate and decide upon the tone of the writing.
The teacher might ask a pupil to come up and write the first sentence and then ask other pupils to comment on it. She might drip-feed technical vocabulary into the conversation where appropriate (reminding pupils, for example, that placing two contrasting ideas side-by-side is called “juxtaposition”) and she might encourage pupils to repeat it and use the correct term in future. She might ask pupils to model their thought processes, thinking aloud as they write, explaining the reasons for their choices.
The teacher, therefore, will mostly be engaged in asking open questions, such as: “Why did you choose that word? “Is there another word which might fit better or have more impact? Why is this word better than this one? Should we use a short sentence here? Why/why not? What is the effect of this, do you think?”
Once the class has constructed a model together, they need to do so independently.
Independent practice not only provides a crucial third opportunity for pupils to practise – after teacher modelling and co-construction – it also enables pupils to demonstrate their own understanding and for the teacher to assess the extent to which they have “got it”. Until a pupil completes a task by themselves, we – and perhaps they – cannot be certain that they can do it or that information has been encoded in long-term memory.
If pupils succeed, the teacher can move on. If not, the teacher can use the feedback information to guide further teaching of the subject, perhaps re-teaching key elements of it or engaging those pupils who have succeeded in teaching those who have not.
The feedback loop
The four-part teaching sequence is not the end of the learning process, because once pupils have practised new learning we need to provide planned opportunities for them to be assessed (by themselves, by each other, or by us) and receive feedback on what they have mastered and what they still need to practise. Then, crucially, we need to provide planned opportunities in class for them to act upon that feedback.
Failure is the best teacher. Pupils learn through practice, by making mistakes, and by experimenting. They also learn best when engaged in a process of trial and error and when they repeat actions several times, making incremental improvements each time.
If we do not provide lesson time for pupils to respond to feedback and improve their work, we send a negative message about the importance of redrafting work and learning from our mistakes. What’s more, if pupils do not respond to feedback in class, the teacher cannot see progress being made and cannot, therefore, recognise and celebrate it.
So what feedback works best?
Feedback needs to motivate pupils to make progress. In this regard, short verbal feedback is often more motivational than long written comments on pupils’ work. Indeed, some pupils find written comments demotivating because they ruin the presentation of their work, are confusing or overwhelming. Feedback should also prompt further thinking and drafting, perhaps by posing questions on which the pupil has to ruminate and act, as opposed to ready-made suggestions and solutions.
I think it is useful to remember (as I explained in my recent series on feedback: http://bit.ly/2yqimEL) that the term “feedback” originated from the field of engineering where it formed part of a loop: feedback in engineering terms was about the discrepancy between the current state and the desired state of something, but this alone was deemed useless unless there was also a mechanism within the feedback loop to bring the current state closer to the desired state. In other words, feedback was about correction and progress.
As such, simply telling pupils that their current performance falls short of where they need to be isn’t feedback in the original engineering sense of the term. Rather, to be effective, feedback must also embody a mode of progression for pupils.
The best feedback causes thinking. In practice, this means that the teacher should be clear and constructive about pupils’ weaknesses, offering suggestions on how they might be addressed, identify pupils’ strengths and offer advice on how to develop them, and then – crucially – provide planned opportunities in class for pupils to improve upon their work.
And that is the four-part teaching sequence I recommend you follow, no matter what you’re teaching and to whom. Tell pupils what you need them to know, show them what it looks like and how it works, produce a model of excellence together, plan opportunities for pupils to practise their learning independently, then make sure practice leads to quality feedback that pupils act upon in class so that progress becomes visible for all and the importance of learning from mistakes is writ large.
Follow me on Twitter: @mj_bromley