The question critics of explicit teaching cannot answer

Try it. Ask them.

Image via AI

---

In 2022, Lin Zhang and colleagues published a paper arguing that recommendations about science teaching in the U.S. were too narrowly focused on one form of research: program-based studies. In these studies, a new, shiny, packaged intervention is compared with business as usual and found to be more effective. The problem with these studies is that they vary more than one thing at a time and because they are not ‘blinded’—teachers and students know whether they are getting the cool new thing or not—we cannot know what role expectations play and therefore cannot rule out a placebo-like effect. Instead, Zhang argued that recommendations should also take account of smaller, controlled educational psychology studies where only one thing is varied at a time and where both conditions are a bit different to normal, as well as large-scale correlational studies, such as the evidence coming from the Programme for International Student Assessment (PISA).

On this basis, Zhang et al. argued the evidence for explicit teaching had been neglected and the evidence for inquiry learning, which is often a feature of program-based studies, had been oversold.

Zhang’s paper sparked a high-profile response from Ton de Jong and twelve of his colleagues that critics of explicit teaching like to highlight. De Jong et al. largely explain away the point about the evidence base that Zhang et al. raised and focus more on the relative merits of explicit teaching versus inquiry learning. On the one hand, they assert the superiority of inquiry learning—“inquiry-based instruction produces better overall results for acquiring conceptual knowledge than does direct instruction”—while also suggesting there is a time for explicit teaching and a time for inquiry learning. You may have noticed people making this exact point online.

I was part of a team with John Sweller, Lin Zhang, Paul Kirschner and William Cobern who then wrote a response to this paper. We offered a definition of explicit teaching:

“To us, the defining feature of explicit instruction is that, for novice learners, concepts are fully explained, and procedures are fully modelled before learners are asked to apply those concepts or procedures. Significantly, this working definition does not preclude the possibility of learners completing open-ended problem-solving tasks. It simply suggests that if such tasks form part of a sequence of explicit instruction, they would follow explanation and modelling by the instructor.”

This links the micro-definition of explicit teaching—fully modelling and fully explaining—with the more expansive one i.e. a whole process that starts with teacher modelling and explanation but gradually hands control over to students so that, eventually, they are solving complex problems and producing complex products on their own. This form of instruction is backed by cognitive load theory, the experiments that support the theory and the concepts the theory puts forward to explain these results, such as the idea of element interactivity.

There are other theoretical frameworks to understand explicit teaching, but cognitive load theory is a particularly fruitful one, offering relevant guidance to teachers at each stage of the learning process. Critically, it provides guidance for when teachers might choose to be more directive or for when they might instead ask students to engage in more open-ended problem solving.

The key point is that de Jong et al. don’t do that. Although they assert that there is a time and place for both explicit teaching and inquiry, they give us no criteria to decide which one to use in any given situation. Given their simultaneous assertions that inquiry is superior for acquiring conceptual knowledge, this is a confusing proposition.

As an aside, although the idea of conceptual knowledge or, more simply, understanding, is appealing, when we examine how researchers try to measure it in experiments, it falls apart and can often look a lot like remembering definitions. In essence, demonstrating this ‘conceptual knowledge’ can be pretty low level compared with demonstrating procedural knowledge in an area like maths or science. This is why, in my PhD research, I chose to use the ability to transfer problem-solving strategies to different problem types as my measure of understanding.

To summarise, de Jong et al. both argue for the superiority of inquiry learning and that explicit teaching and inquiry learning should be used at different times, without specifying criteria for deciding which to use.

We have seen exactly this argument before. When whole language reading instruction was shown to be ineffective, both through research and problems with its wide-scale adoption in places like California, advocates hit on a new idea. ‘Balanced literacy’ was born. This was supposed to blend the best of whole language and explicit literacy teaching but, similar to de Jong’s balance between explicit teaching and inquiry, it was never clear when or why a teacher should use one strategy over the other.

The programs that came out of balanced literacy, such as Lucy Calkins’ Units of Study and Fountas and Pinnell’s approach, did not align with the evidence from reading research, as Emily Hanford has pursued through her reporting and her Sold a Story podcast. This has led to a backlash against these programs and the tumbling of the edifice of balanced literacy.

Nevertheless, the idea of striking some balance between two positions that are, by implication, extreme, sounds reasonable, has a great deal of appeal and worked extremely well for proponents right up until the point that it didn’t.

It may be that de Jong et al.’s formulation—balanced teaching—has similar appeal.

Balanced teaching is essentially atheoretical. To the extent it is supported by any coherent theory, it is mushy constructivist teaching theory and this only applies to the inquiry learning half of the formulation. Therefore, the killer question to ask advocates is precisely when they would use explicit teaching and precisely when they would use inquiry.

You won’t receive a coherent answer because they don’t know. At best, you will be offered a word salad. At worst, you will be impugned as too stupid or lacking in nuance to understand.

Nevertheless, if you get the chance, give it a go and see if I am right.

Finally, I should add that there was a last paper in this exchange. Educational Research Review asked de Jong and colleagues to write a response to our response. You can find it here and it is so bad it is essentially self-refuting.