Stretch and challenge? Only for the few

Posted on 03/06/2026

A recently-published EEF study of student grouping for maths lessons prompted misinformed headlines and buttressed the ‘ability’ discourse that so damages education. Patrick Yarker’s blog about the study follows those by Rachel Marks and Perpetua Kirby

How good are you at spotting differences?

The mixed attainment group [of schools] had an average KS [key stage] 2 intake score of around 29 points. Around a third of schools in this group were rated Ofsted Outstanding and 15% Requires Improvement. FSM [free school meals] and EAL [English as an additional language] proportions were 22% and 20%, respectively. The majority of schools fell within IDACI [income deprivation affecting children index] quintiles 1 – 3, and 81% were urban.

The setted group had an average KS2 intake scores of around 29 points. Over a third of schools in this group were rated Ofsted Outstanding and 3% Requires Improvement. FSM and EAL were 19% and 14%, respectively. The majority of schools fell within IDACI quintiles 1 – 3, and 90% were urban.

These two paragraphs describe the characteristics of two groups of schools involved in a recent Education Endowment Fund (EEF) research project.[1] The study ‘compared the attainment and self-confidence of Year 7 and Year 8 pupils (aged 11 – 13) taught in mixed-attainment [maths] classes with those taught in [maths] classes set by prior attainment across 97 schools’.

The findings of the project were trumpeted in the national press. They purported to show that mixed-attainment classes are detrimental only to the students in them with high prior attainment. These students made ‘two months less progress’ than did students in classes stratified by attainment – presumably top sets. I wanted to know which two months: the full measure of January and March, say, or the short change of February and June. But the study does not say.

One justification offered by EEF for spending almost £1 million on this project was that it provided ‘up-to-date evidence’. The EEF website claims that ‘there is limited evidence on the impact of different grouping practices on pupil attainment within the English education system’. This claim is spurious. There is a mountain of evidence going back decades. The EEF chooses to overlook much of it, supposedly on methodological grounds. But at root this is a political decision, not a scientific one. It is a choice to ignore, and by implication delegitimise, the significance of what people adversely affected by ‘ability’ labelling and grouping have to say about their experiences in many a study. As a form of not listening, the EEF researchers’ stance is compounded by the lack of citation in its report for significant work by scholars such as Susan Hart, Annabelle Dixon, Mary Jane Drummond, Donald McIntyre, Mandy Swann, Holly Linklater, Jo Boaler, Susan Hallam, Samantha Parson and Rachel Marks. Their work has enhanced our knowledge and understanding of the enduring damage done by fixed ‘ability’ labelling, thinking and practice.

The EEF describes its project as adopting ‘a quasi-experimental design’ which ‘matched the two groups of schools to ensure they were as similar as possible across key attainment characteristics’. A look back at the two paragraphs presented above for comparison renders this assurance of similarity hard to credit. The proportion of schools requiring improvement in one group was five times as large as in the other group (15% against 3%). The proportion of students in receipt of free school meals was more than 10% higher in one group than in the other (22% against 19%). And the proportion of students for whom English is not a first language was almost 50% higher in one group than the other (20% against 14%).

These characteristics speak to the being of each school as an institution. Discretely and in combination they will have a shaping and lasting influence on what goes on there. The variations underscore how problematic and doubtful are the grounds on which one school may be compared with another, before the difficulties are multiplied many times over when it comes to comparing classes or groups of students. To describe the research as ‘quasi-experimental’ sounds like a desire on someone’s part to co-opt the prestige of science while not being actually mendacious.

The study uses the language of ‘prior attainment’. But there’s no escaping the way such language plays into the ‘ability’ discourse which pervades the school system and is used to justify the grouping of students into supposedly homogenous sets, groups or tables. The reporting of the project, for example in the Guardian, veered back and forth between the terms ‘attainment’ and ‘ability’. Within the pervasive ‘ability’ discourse, attainment functions as one of the proxies for ‘ability’ (itself a euphemism for ‘intelligence’) and by which is meant fixed innate ability. The EEF project report itself taps into ‘ability’ tropes, for example in its recommendation that: ‘Clear strategies for stretching and challenging high attainers should be embedded as a core part of lesson planning’. This assumes that ‘high attainers’ are straight-forwardly distinguishable from those who are ‘middle attainers’ or ‘low attainers’: like these others a reified and discrete group. What makes them so? Not their ‘high attainment’. That simply signals their high ‘ability’ at maths, and it is ‘ability’ that is underneath the attitude presented here. An attitude which, apparently, sees no need to have strategies to ‘stretch and challenge’ all attainers. Why offer ‘low attainers’ stretch and challenge? After all (according to this way of thinking) they will never be ‘high attainers’. And in an education system driven by results, that is what matters.

Ostensibly independent of government, the EEF aligns itself with the Department for Education’s declared areas of research interest. It has been given some £260 million of public funds. Research it produces is primed to carry weight among policymakers keen to push efficiency and (in their terms) effectiveness. This latest EEF project services a way of thinking about and working with pupils and students which upholds notions of fixed ‘ability’ and endorses policies which segregate, stratify and exclude.

The damage to pupils’ and students’ self-image, self-esteem, engagement and motivation which ‘ability’ labelling and grouping inflicts has been often rehearsed but rarely heeded. They suffer from it who are labelled high ‘ability’ as well as ‘low’. The way of thinking which, in the EEF manner, can approach educational growth and development – call it ‘progress’ if you like – as something available to be quantified on a linear scale, and with the kind of exactitude which talks about ‘two months’ worth’ of difference, is more than impoverished. It is poison to the system. Every day in school is blighted by it.

Learning is more like coming to know a country than climbing a ladder. Thus Lawrence Stenhouse, whose centenary falls this year. For the EEF researchers he couldn’t be more wrong. Learning, at least as the EEF would construct it, is less and less like coming to know a country and more and more like having to climb a ladder. Or – thanks to the ‘ability’ discourse – being prevented from climbing it.

Further shortcomings of the EEF are anatomised by Rachel France in the current issue of FORUM. Rachel Mark’s blog about the EEF grouping study is here; Perpetua Kirby’s is here.

Patrick Yarker

June 2026

[1] The paragraphs, and all other quotations, can be found on the EEF website: https://educationendowmentfoundation.org.uk/projects-and-evaluation/projects/student-grouping-study