21st September 2023

Variation in pupil inclusion practices across schools: findings from the second wave of the DEEP survey

This blog examines the results from the second wave of the Decisions in Education in England Panel (DEEP) survey, which is part of a multi-year mixed methods project which aims to identify the most effective MATs and school groups (including local authorities, dioceses and federations), supported by GL Assessment.


As of August 2023, there were 2,397 academy trusts, consisting of just over 10,000 schools operating in England, numbers which are steadily increasing.[i] The rising number and size of multi-academy trusts (MATs), has led to calls for greater transparency and for MATs to be held accountable across a wider range of outcomes. [ii] Previous research has primarily examined MATs’ performance using attainment indictors. For example, we have previously shown that MATs are over-represented amongst the lowest performing school groups based on GCSE results, although there was substantial variability within and between school groups.[iii] Whilst these and other findings provide commentary on performance less work has focused on understanding how MATs and other school groups are performing on measures beyond attainment, such as attendance, inclusion, and wellbeing.

In light of this, we launched a national survey, the Decisions in Education in England Panel (DEEP) survey, to gain a more detailed and nuanced picture of what decisions, actions, and policies are currently being implemented in schools. This survey forms one part of a larger multi-year mixed methods project which aims to understand the key features of effective school groups. The project also consists of the development of a series of quantitative measures using administrative data. Our project is designed to give school leaders and policymakers a more robust understanding of how the most effective trusts and school groups operate.

In this second wave of the DEEP survey, we sought to obtain a more comprehensive understanding of how schools use admissions, teaching, and behaviour management policies and practices to improve pupil inclusion and outcomes. 239 school leaders and 86 school group leaders were surveyed of which 96% worked in a MAT. Drawing on the results from the second wave, this blog presents three charts that provide some insights into schools’ current inclusion practices. As the DEEP survey forms only one part of a larger project, this blog does not seek to provide recommendations regarding best practice, but instead provide an overview of how and what policies and practices are currently being implemented in schools. We will be using this analysis in future work to help triangulate the findings from our quantitative measures and determine the attributes of an effective school group.  


1. Segregation by attainment increases as pupils progress through education


Findings from the DEEP survey indicate that pupils are typically further stratified as they progress through education but that within school phase there is variation in the extent grouping by prior attainment is used. We categorised grouping by prior attainment into 4 groups: mixed ability grouping, setting, streaming, and within class grouping. Setting refers to the practice where pupils are grouped into classes by ability in specific subjects; streaming/banding, where pupils are grouped into the same classes by ability for most of their lessons; and within-class grouping where pupils in mixed-ability classes work in groups of similar ability/attainment. To produce the above chart, we asked school leaders to ‘Please select which methods of ability grouping are used by different key stages in your school.’

Upon entering school in reception, pupils are most likely not to be grouped by their attainment with roughly 90% of sampled school leaders reporting they use mixed ability grouping. Throughout primary school the use of attainment grouping increases, particularly within class grouping, although mixed ability grouping remains the most common practice. We observe a sharp change in practices in secondary schools – the use of mixed ability groups falls and the use of setting, increases. Almost all sampled schools use some form of setting and/or streaming at key stage 4.

The effects of attainment grouping on performance is unclear. From the limited existing research into attainment grouping, findings indicate that setting and streaming may have negative social and academic effects for pupils placed in lower sets, and these pupils tend to be disproportionately from socio-economically disadvantaged backgrounds.[i] Research has also found that pupils placed in higher sets tend to benefit from higher rates of attainment, potentially exacerbating inequality in the classroom.[ii]


2. Schools prioritise admitting pupils based on siblings and proximity instead of need[1]


Schools in England are able to set their own oversubscription admissions criteria, provided they are consistent with the School Admissions Code.[i] The surveyed schools, both primary and secondary, typically give priority to pupils that already have siblings at the school and those living in close proximity. That said, roughly a quarter of primary schools also appear to prioritise pupils with complex needs. This contrasts with secondary schools of which only 10% seem to prioritise pupils with complex needs. Instead, priority tends to be further directed towards children of staff and pupils from feeder schools. Despite such differences between school phases, siblings and proximity are the most widely used admissions criteria in the event of oversubscription within our sample.

Prioritising based on family and distance from the school likely arises for practical reasons and possibly in efforts to appeal to the local community and parents rather than inclusion or equity. However, we know that residential segregation means distance is not neutral and families with more purchasing power may be able to more readily exercise school choice as they can chose to live closer to desirable schools.[ii] In the quantitative strand of our work programme we hope to further explore how representative schools’ intakes are of their local communities, and if this is affected by admissions policies.



3. Internal exclusion provision is widespread and varied in schools[1]


Unlike a fixed-term exclusion, where a pupil is sent home and responsibility is placed on parents/carers, an internal exclusion means a pupil remains in school but is removed from their normal classes. There is no formal prescription regarding its usage and there are no national statistics confirming the number of schools using this practise, how long pupils are typically internally excluded for, or what pupils are doing whilst excluded from the classroom. The typical rationale for using this policy is that exclusion does not function as a reward (‘time off school’), as excluded students continue to receive education and adequate supervision in a more managed environment.

The findings from the DEEP survey indicate the use of internal exclusion is more prevalent in secondary schools – less than 3 per cent of sampled secondary schools reported not using internal exclusion at all, in comparison with almost a quarter of primary schools. Secondary schools are also most likely to send excluded pupils to designated ‘isolation exclusion units’, whereas in primary schools, the most common form of internal exclusion is the senior leader’s office. 65% of primary schools reporting using senior leaders’ offices as a form of internal exclusion. Only a small number of schools do not use any form of internal exclusion, the majority of which are primary schools. For schools that have the option to use internal exclusion, there is considerable variation in its provision. Such diversity in internal exclusion provision may be promising as behaviour management is tailored to pupils’ specific needs. However, it may also indicate a lack of clarity regarding best practice. Furthermore, the survey is unable to unpick exactly what educational provision excluded pupils are able to access in each of these different settings.



In conclusion, findings from wave 2 of the DEEP survey indicate that pupils are increasingly grouped by attainment as they progress through schooling, admissions priority is typically directed towards pupils with siblings at the school and those living in close proximity rather than by need, and that the use of internal exclusion is widespread and varied particularly amongst secondary schools. Our findings also demonstrate there is currently considerable variation across schools and emphasise the need to understand and critically question policies and practices utilised within schools and how they may influence outcomes. Through our future development of a robust series of quantitative measures alongside the DEEP survey, we aim to begin to address these gaps in knowledge by examining what an effective school group looks like.

[1] School leaders were asked to ‘Please indicate the order in which the following criteria give priority to pupils when admissions are oversubscribed AFTER pupils with EHCP plans which name the school and AFTER looked after children and previously looked after children.’

[2] This chart was created based on the following survey question: ‘What form of internal exclusion is used in your school, if any?’

[i] Department for Education, ‘Open Academies, Free Schools, Studio Schools and UTCs’, GOV.UK, 2 May 2023, https://www.gov.uk/government/publications/open-academies-and-academy-projects-in-development.

[ii] Nerys Roberts, ‘March 2022 Schools White Paper (England)’, 16 August 2023, https://commonslibrary.parliament.uk/research-briefings/cbp-9511/; Anne West, ‘High Stakes Testing, Accountability, Incentives and Consequences in English Schools’, Policy & Politics 38, no. 1 (2010): 23–39, https://doi.org/10.1332/030557309X445591.

[iii] Jon Andrews and Natalie Perera, ‘The Impact of Academies on Educational Outcomes’ (The Education Policy Institute, July 2017), https://epi.org.uk/publications-and-research/impact-academies-educational-outcomes/; Jon Andrews, ‘School Performance in Multi-Academy Trusts and Local Authorities – 2015’ (Education Policy Institute, July 2016).

[iv] Diane Reay, ‘THE ZOMBIE STALKING ENGLISH SCHOOLS: SOCIAL CLASS AND EDUCATIONAL INEQUALITY’, British Journal of Educational Studies 54, no. 3 (1 September 2006): 288–307, https://doi.org/10.1111/j.1467-8527.2006.00351.x; B Francis, B Taylor, and A Tereshchenko, Reassessing ‘Ability’ Grouping: Improving Practice for Equity and Attainment, 1st ed., 2019, https://doi.org/10.4324/9780429436512; Adam Gamoran, ‘Tracking and Inequality: New Directions for Research and Practice’, in The Routledge International Handbook of the Sociology of Education, 1st ed., 2010; Judith Ireson, Susan Hallam, and Clare Hurley, ‘What Are the Effects of Ability Grouping on GCSE Attainment?’, British Educational Research Journal 31, no. 4 (1 August 2005): 443–58, https://doi.org/10.1080/01411920500148663.

[v] ‘Setting and Streaming’, Education Endowment Foundation, accessed 29 August 2023, https://educationendowmentfoundation.org.uk/education-evidence/teaching-learning-toolkit/setting-and-streaming.

[vi] Department for Education., ‘School Admissions Code’, GOV.UK, 11 March 2022, https://www.gov.uk/government/publications/school-admissions-code–2.

[vii] Amanda U. Potterton et al., ‘Sociological Contributions to School Choice Policy and Politics Around the Globe: Introduction to the 2020 PEA Yearbook’, Educational Policy 34, no. 1 (1 January 2020): 3–20, https://doi.org/10.1177/0895904819881150; Carol Vincent, Annette Braun, and Stephen Ball, ‘Local Links, Local Knowledge: Choosing Care Settings and Schools’, British Educational Research Journal 36, no. 2 (1 April 2010): 279–98, https://doi.org/10.1080/01411920902919240.

GL Assessment’s goal is to enable every child to realise their full potential, using expertise in assessment gained over decades of working with schools and teachers. GL Assessment takes a ‘whole pupil view’ which takes into account ability to achieve, current attainment, and any barriers to learning they may have 

They work with partners, including King’s College London and the University of York, to ensure that assessments are their assessments rigorous, academically sound and in line with current best practice in education. Their assessments are also widely used by the Education Endowment Foundation (EEF) to measure the impact of their intervention research, and they are also used by schools and Governments in over 100 countries worldwide. GL Assessment are proud to be part of Renaissance, a global leader in education technology.