If it’s not child-centred, who’s it for?

England’s ‘Official Pedagogy’ – and the evidence it left out

I find it tedious to put away the washing, so I invented something called ‘pants party’. Everyone sorts out their own washing while we listen to music and the children dance. The marketing is doing a lot of heavy lifting here.

Last night, Blossom¹ (8) chose a song she’d heard that day at school – ‘The Vertebrates Song’, by Hopscotch.

Repeat after me: Mammals, reptiles, birds, amphibians, and fish.

Rosie (4) asked Blossom what amphibians are.

‘Frogs and toads,’ Blossom said. Then turned to me: ‘And newts too?’

I hedged my bets. They live in ponds, and they’re a bit slimy. The sort of thing you really need a science teacher on hand for.

Animal classification is something we teach in school – well, science teachers do. It’s part of the powerful knowledge our curriculum is built on. And according to the orthodoxy that now governs English classrooms, that knowledge belongs in school: carefully sequenced, taught in small chunks to avoid cognitive overload, with pupils securely holding the knowledge before any enquiry begins, moving from novice towards expert.

So what was Rosie doing asking about amphibians when she had only half an idea about animal families? After nursery today she asked why butterflies visit flowers as well as bees. Sometimes she’ll ask question after question, blissfully unaware she’s firmly in the novice camp.

And there’s nothing special about Rosie, beyond her love of clothing emblazoned with unicorns and her vehement dislike of anything containing vitamins or fibre. Questioning is something all kids do. No sooner have they learnt the word ‘why’ than they’ve weaponised it.

When children start school – as everyone from Bourdieu to Hirsch agrees – some arrive with more of this powerful knowledge than others. Despite the best evidence-backed methods we have, our schools struggle to close that gap.

At age 11, less than half of disadvantaged pupils reach the expected standard in reading, writing and maths combined, compared to more than two-thirds of their wealthier peers. Even if two children score the same at 11, by 16 the disadvantaged child is more than half a grade behind, on average, across all their GCSEs. While 44% of the GCSE gap is already present at 11, the remainder opens up during secondary school.

We spend £3 billion a year on Pupil Premium funding to support some 2.3 million disadvantaged pupils, and the Education Endowment Foundation has been given hundreds of millions to get at the causes.

We focus on the outcome of what happens at home, the visible bit: the fact that some kids arrive at school with more powerful knowledge than others. What we should be asking is how they got it in the first place.

Blowing up the concrete

No politicians in recent decades have shaped English education more than Nick Gibb and Michael Gove. Gibb’s holiday reading in 2005, when he was shadow Minister for Schools, was E.D. Hirsch’s The Schools We Need and Why We Don’t Have Them. Hirsch argues that educated citizens must be fluent in shared cultural knowledge; where the home hasn’t supplied it, the school must. Gibb recommended it to Gove, and when the Conservatives took power, England’s knowledge-rich curriculum was born.

Gibb had an enemy in mind. ‘I am not sure until Michael and I came in that people really understood that the problem with poor standards is progressive education ideology, the child-centred learning approach to teaching,’ he later said.² Drastic action was required: ‘In 2010… we had to blow up the concrete to allow teachers to have these [opposing] views’.

Rightly, Gibb insisted pedagogy should be backed by evidence: ‘Poor teaching methods harm all pupils, but a growing body of research suggests it harms disadvantaged pupils most of all’. But this is not just about two men from one party. The vision survived ten education secretaries and a change of government without Labour laying a finger on it – not least the teacher training framework, which now sets the standard for good teaching in England.

England’s ‘official pedagogy’

In 2019, for the first time, England’s teacher training providers had a curriculum of their own to follow, citing familiar names: Wiliam, Dunlosky, Willingham, Sweller, Kirschner, Hattie, Rosenshine.³ It hasn’t been universally popular, especially among academics, two of whom argued it produces an ‘official pedagogy’ resting on ‘an imaginary notion of instructional practice that does not fully grasp the context of teachers’ work’.

‘Explicitly teaching pupils the knowledge and skills they need to succeed within particular subject areas is beneficial,’ teachers are told, and they should do it by ‘breaking complex material into smaller steps’.

The framework doesn’t represent poor practice. But it does, I’ll argue, represent a partial view of the evidence on learning – one that, followed to the letter day in and day out, can clash with the best evidence we have about using children’s curiosity to drive their learning. Effective teaching can certainly motivate: you just have to watch the realisation cross the face of a struggling student when an idea finally clicks. But it isn’t the only way, as the evidence on out-of-school learning shows.

Young children learning

A few years ago, when Blossom was five or six, her favourite book was one I was using for my PhD – Tizard and Hughes’ Young Children Learning. It’s a 1984 study of 30 four-year-olds, tape recorded at home and at preschool. We’d act out the transcripts; she’d be the daughter, I’d be the mother.

This is what Tizard and Hughes concluded:

Our observations of children at home showed them displaying a range of interests and linguistic skills which enabled them to be powerful learners. Yet observations of the same children at school showed a fundamental lack of awareness by the nursery staff of these skills and interests… in the world of school, the child appears to be a much less active thinker than is the case at home.

It wasn’t meant as a criticism of staff, but a point about the quality of interaction between a child and someone who knows them well, versus someone supervising a dozen others. Later work supports it: Rachel Romeo and colleagues found that – after controlling for IQ, socioeconomic status and sheer volume of speech – it was the number of conversational turns, the back and forth between child and adult, that predicted a child’s verbal skill.

Opponents like Gibb often associate child-centred learning with leaving children to their own devices. These studies show the opposite: it requires a deep understanding of the child, expressed through responsive interaction – a form of scaffolding. The knowledge gap may itself be partly a product of fewer such interactions, on average, in disadvantaged homes.

It’s also unlikely much of that back and forth is the cognitive-load-respecting direct instruction the framework prizes. It’s hard to imagine many dinner table conversations like this:

Right, before pudding, tell me the chemical composition of water, little Bethan. Two elements, yes. No – let’s try another and come back. This is CO₂; the absence of a number after the C indicates one element of… what was the C? Yes, carbon. Now, the O…

There’s also evidence that direct instruction can narrow children’s exploration. Elizabeth Bonawitz and colleagues gave preschoolers a novel multifunction toy; children explored its other features less if an experimenter explained one – an effect that held even for children who’d merely overheard the explanation. Those given no explanation explored more widely.

None of this means direct instruction is ineffective, or that parents never explain things at length (although direct instruction involves more than just this). But it’s a mistake to assume there’s no robust science behind child-centred methods. In fact, we’re just getting started.

A small step for knowledge – or should that be a giant leap?

England’s teacher training framework presses teachers to break things into small steps and avoid overloading working memory. This sounds innocuous, but to the emerging science of curiosity, this advice could pose a threat.

George Loewenstein, in his seminal 1994 paper, The Psychology of Curiosity, claimed that curiosity is triggered by a knowledge gap of just the right size – something that has now been shown across many studies. Curiosity researchers don’t treat ignorance as something to eliminate as fast as possible. They treat a carefully judged gap – uncertainty of the right size – as the engine of learning itself.

The framework encourages teachers to minimise unnecessary uncertainty by ensuring pupils master the knowledge before moving on, to ‘avoid overloading working memory’. But curiosity doesn’t want small steps; it wants a gap tailored to the learner’s confidence, and it doesn’t want that gap closed immediately – it wants you to sit with it. That’s curiosity, after all. The framework, by contrast, tells teachers to ensure pupils ‘master foundational concepts and knowledge before moving on’ and to ‘remove scaffolding only when pupils are achieving a high degree of success’. There’s a central message: knowledge first.

Manu Kapur’s work on productive failure turns that on its head. When learners attempt something and don’t succeed – and are then taught the key points – they often understand the underlying concept better and transfer it more readily. Direct instruction still features; it’s the order and format that change.

We can see productive failure as curiosity engineered into a lesson: the problem-first element forces pupils to draw on prior knowledge, and the gap they hit can activate curiosity – provided it’s well judged and students value the information required to fill the gap. That’s where responsive teaching comes in: scaffolding to size the gap, and care in choosing the problem so the payoff feels worth it.

There are some reasonable criticisms – failure can demotivate, which is precisely why scaffolding matters. Productive failure and cognitive load theory needn’t be at odds, though. With scaffolding to manage load (or distributed across a group of learners), the problem-first phase needn’t overwhelm working memory.

The problem with adaptive teaching

Nick Gibb was opposed to child-centred methods – so how then do you support disadvantaged pupils and those with special educational needs? The framework’s answer is an awkward mix of compulsion and proscription. Teachers must grasp that ‘pupils are likely to learn at different rates and to require different levels and types of support’ – blandly obvious – except that in practice strategies are ruled out: adaptive teaching is ‘less likely to be valuable if it causes the teacher to artificially create distinct tasks for different groups of pupils or to set lower expectations’.

There’s evidence from the EEF – who helped write the framework – that brings this advice into question. Across 198 studies (admittedly of varying quality), their toolkit rates individualised instruction – ‘providing different tasks for each learner and support at the individual level’ – as having ‘moderate impact for very low cost,’ and says, for pupils who have fallen behind, it ‘may allow the teacher to provide activities that are closely matched to a pupil’s attainment’.

There’s little evidence the current methods are working for the learners who most need support. England is facing what parliament calls a ‘SEND crisis’: 1.7 million pupils with SEN, ‘exhausted parents fighting for basic support, teachers stretched beyond capacity’, and a system that the House of Commons committee concluded unambiguously ‘is not working’.

It’s little better for disadvantaged pupils. The EEF was set up to find ‘what works’ to close the gap, especially for those eligible for free school meals. Despite promising findings, we haven’t found it: across the trials they studied, Bilal Ashraf and colleagues found the equivalent of 1 month’s progress for literacy and no effect at all for maths among pupils on free school meals. And even that flatters whole-class approaches.

Across large-scale evaluations, the strongest effects came from targeted one-to-one and small-group literacy interventions. When aggregated, whole-class and whole-school interventions would be classed as having ‘very low or no impact’ for disadvantaged students.

Tailoring is the common thread. A teacher working with one child or a handful of students can gauge the perfect knowledge gap for the learner in front of them – exactly the responsiveness that curiosity research points to, and exactly what whole-class delivery of fixed small steps makes hardest.

Breaking out of the straitjacket

Hirsch, Gibb, Gove and the other advocates of a knowledge-rich curriculum began with an admirable goal: to help the most disadvantaged. They saw a product – the knowledge some children already have when they arrive – and set out to make up the deficit in those who didn’t.

What they underestimated was the mechanism.

Disadvantaged children are as curious as wealthier ones; the difference, as my analysis of the OECD’s PISA data shows, is their love of learning in school. Poorer pupils who keep that love do better, but my analysis shows that England has the widest gap out of 75 education systems between general curiosity and that school-specific love of learning for the wealthier and the disadvantaged.

We’ve built an official pedagogy with a great deal to say about what children should know, and surprisingly little about how to keep them wanting to know it. We can fix that – keeping the most powerful parts of explicit teaching while experimenting with those knowledge gaps until they spark curiosity. It may mean straying beyond the bounds of the knowledge-rich curriculum, but should that really scare us?

The curriculum was always meant as an entitlement.

Why, then, have we let it become a straitjacket?

1. All participants selected pseudonyms. Permission sought from parent (me) prior to publication.

2. Rousseau is often named as the father of the child-centred approach, which is a little ironic given he left all five of his children at the local foundling home (he did later regret not making a note of their dates of birth or gender, at least). I don’t think this renders his ideas – or the wider philosophy – irrelevant, though I agree it doesn’t make him the ideal exemplar.

3. The prototype was proposed in 2016. The 2019 framework was updated in 2024, but the underlying pedagogical model and its evidence base barely changed.