Grouping by attainment versus "mixed-ability" grouping remains one of the most hotly contested issues in schooling, both for researchers and teachers.
The vast majority of education systems segregate pupils by attainment in various ways – whether by practising "between school" segregation (as with our prior tripartite system and remaining grammars) or "within school" attainment grouping (practices of streaming, setting or within-class grouping such as "ability tables").
The English education system is relatively comprehensive in that the vast majority of schools are open to all, irrespective of attainment, and retain a comprehensive curriculum to age 16.
However, the Organisation for Economic Cooperation and Development found it to have unusually high levels of within-school grouping practice.
So does this matter?
The research literature in this area suggests that the application of attainment grouping does not, overall, have a significant effect on pupil outcomes in comparison to heterogeneous grouping.
But, within this, there is evidence that attainment grouping has a negative impact on the outcomes of low attainers – and some evidence of a positive impact on the highest attaining pupils (those often referred to as "gifted" in the British system).
The impact of setting on social justice
This, in turn, has implications for social justice, because pupils from low socioeconomic backgrounds – and some other groups such as black boys – are over-represented in low sets and streams.
As such, they are arguably subject to double disadvantage: they are disadvantaged on arrival into the education system, and then they are the recipients of school practices found to hinder their progress.
We were surprised at the lack of specificity of much of the existing research, meaning that teachers would have little idea what was causing these apparent effects, and how practice might consequently be improved.
Especially, the research mainly fails to disaggregate the different explanations for differential impacts of attainment grouping on different groups.
Or, indeed, look at why low attainers appear to perform better in heterogeneous groups – the general lack of research on the constitution of good practice in mixed attainment grouping is striking.
Our research, funded by the Education Endowment Foundation, set out to address these issues.
It involved randomised controlled trials of two interventions.
The first, a fully powered study involving 126 secondary schools, explored what we called "Best Practice in Setting" – setting practice following a set of research-based principles aimed at mitigating negative practices associated with setting.
This sought to test whether the intervention could improve outcomes for pupils (especially those in low attainment groups).
The second, a small feasibility study involving 13 secondary schools, explored "Best Practice in Mixed Attainment Grouping", seeking to establish principles.
What this research doesn’t do is offer a comparison of setting and mixed-attainment grouping, so we cannot draw conclusions about which is more effective.
As it was, the interventions found no significant effects.
Given that the two-year intervention suffered from challenges to retention and regarding the extent to which the intervention was implemented as intended ("fidelity"), the findings on the effects of the interventions have limited security.
Nevertheless, the wider project findings present an array of evidence to teachers that can improve our understanding and practice going forward.
Here are just a few findings on setting:
- Patterns of allocation to sets mirrored the patterns identified in prior research, with more free school meals and disadvantaged pupils in low sets, and the reverse in higher sets; more black students in low sets; and more boys in low sets for English, and girls in low sets for maths.
- One third of pupils were misallocated to a set level, according to their key stage 2 maths results. And of these, black pupils were 2.5 times more likely than white pupils to be misallocated to a lower set. Girls were more likely to be misallocated to lower sets in maths, and boys in English. We did not find evidence of misallocation on the basis of social background.
- Timetabling and class size often determine set placement, rather than pupil attainment.
- Teachers highly qualified in their taught subject were less likely to be allocated to low sets. This pattern was less evident in intervention schools.
- Many pupils felt that teachers taught differently to high and low sets, with higher expectations and standards of behaviour "strictness" applied to high sets. Some low set pupils felt limited and "babied" by their teachers.
- There is a significant correlation between set placement and self-confidence in the set subject, and a correlation between set placement and general self-confidence in learning; suggesting a self-fulfilling prophecy from attainment labelling.
- Our findings suggest that there is a relationship between student liking for school and their attainment set, over and above students’ prior attainment. This relationship is more evident in set placement in mathematics than English.
We are keen to work with teachers to think about improving practice and equity, and have developed some evidence-based tips – our "Do’s and don’ts" – to help.
In the UK, within-school grouping is seen by many as the natural way to do things, yet the evidence indicates that it widens the attainment gap whilst making little overall impact on attainment.
In the absence of developed evidence and resources for teachers on best practice in grouping, generating professional discussion to improve all grouping practices to facilitate good educational outcomes and equitable access to high-quality provision is a vital first step.
Professor Becky Francis, Professor Jeremy Hodgen, Professor Louise Archer, Dr Becky Taylor, Dr Antonina Tereshchenko, UCL Institute of Education
