Rendered invisible

Posted on 20/05/2026

In our second blog highlighting the shortcomings of a recent EEF study on student grouping, Perpetua Kirby shines a light on what ‘attainment-focused’ research doesn’t see.

At secondary school, I was always at the bottom of the top maths set. Whilst I achieved good grades in exams, I never felt particularly proficient or at home in the subject. Setting creates a performative culture, and I was too aware of my marginal status and worked to avoid the precipice and descent into lower sets.

Recent research by the Education Endowment Fund (EEF) heralds the benefits of maths setting in Years 7 and 8 for the attainment of those in top sets, with no apparent costs to students in lower sets. However, this conclusion rests on a narrow conception of educational value and an expansive interpretation of what attainment data can reasonably claim. Drawing on my own research, I suggest that setting, alongside other practices that reinforce meritocratic hierarchies, carries costs that are rendered invisible by attainment‑focused analyses. These include heightened stress, fragile self‑worth, constrained relationships, reduced openness to uncertainty, and diminished opportunities to connect meaningfully to curricular content and, hence, to the world.

I undertook an ethnographic study (2019a; 2019b) in a Year 1 classroom in which children were stratified by attainment in both maths and English. The top maths set was called rectangle because it was seen to be the most complex shape, followed by square, triangle, and a circle – which has no sides – for the bottom set. The highest English set was giant and the lowest beanstalk. Children aged five and six quickly knew their place: ‘[Ours] is the smartest group’, one of them told me.

Performing cleverness

Setting encourages performative ‘cleverness’ rather than deep engagement. Those in the highest groups were mostly male, middle class and among the oldest in their year group: they were frequently told they were clever and asserted themselves as ‘really clever’. Children in the lower sets worked hard to present themselves as clever, but more tentatively (‘I’m a bit clever because I’m still learning’) and self-protectively (‘I was trying to write something, but I was too smart because I am too smart to concentrate’).

The emotional and existential consequences of setting are not necessarily visible to attainment-focused research. By taking time to sit alongside children in the classroom and to speak with them over time, I observed how they demonstrated a willingness and effort to determine and deliver on what is expected – performing cleverness by finding the correct answer – but that they had to negotiate, largely alone, emotions of not knowing what is expected and failure. This could feel frightening – described by one as ‘emergency feelings’ – and did not serve them well when tackling new challenges where there is no clue to the answer.

Even the apparent winners of setting can struggle when asked to engage with uncertainty and judgement – capacities that cannot be scaffolded by clues or correct answers. Those in the top sets, who excelled at demonstrating their knowledge in maths and English tasks, could be thrown by me asking for their opinions. Such open questions require children to deliberate and reflect so as to come to a position that is uniquely their own.

The head teacher of the school felt pressure to use sets despite thinking them ‘less-than-ideal’, given children become ‘really aware’ of their place in the hierarchy. She and her staff were similarly measured, ranked and judged; so too is the government, whose performance is compared on global league tables, notably the Programme for International Student Assessment (PISA). Professional accountability frameworks value what is easily measured and compared. Similarly, being ‘clever’ in school reflects a concern with value and the symbolic power of the quantifiable, evidenced by setting and test scores.

No negative impacts?

The EEF are giants in the world of educational research, with their access to significant funding and their assertions of ‘quasi-scientific’ rigour and scalability. Their study compares maths attainment across different groups, but accounting for the variance in the participating schools and children is complex to achieve meaningfully. The claim by the study’s authors that it shows setting ‘has no negative impact on lower-achievers’ (cited in The Guardian) ignores how statistical relationships offer no such certainties, only possibilities. This statement is a flawed step away from the more cautious assertion on the study’s website that setting ‘does not appear to significantly harm’ attainment for those with low prior attainment  or who are socioeconomically disadvantaged. This shift to certainty sidelines the complexity of scientific method and there remains insufficient discussion of how statistical claims are constructed. The study also neglects wider educational concerns beyond maths attainment.

Even if setting could be conclusively shown to enhance the mathematical attainment of those in the top set without cost to attainment for those in lower sets, this alone is insufficient reason to continue with the practice without far more nuanced deliberation. The educational theorist, Gert Biesta, reminds us ‘that questions about “what works” – that is questions about the effectiveness of educational actions – are always secondary to questions of purpose’, and that ‘values come first.’ The results of the EEF study need to be considered within a wider national conversation about how schools might enhance the possibility for all students to exist and act in the world beyond performing what is measured.

Rachel Marks writes about the EEF Report here.