GCSE Outcomes for Pupils with EHCPs: Mainstream versus Special School Settings. Technical Methodology Report with Statistical Appendices
Saved in:
| Title: | GCSE Outcomes for Pupils with EHCPs: Mainstream versus Special School Settings. Technical Methodology Report with Statistical Appendices |
|---|---|
| Language: | English |
| Authors: | Neide F. Phillips, Department for Education (DfE) (United Kingdom) |
| Source: | UK Department for Education. 2026. |
| Availability: | UK Department for Education. Castle View House East Lane, Runcorn, Cheshire, WA7 2GJ, UK. Tel: +44-37-0000-2288; Fax: +44-19-2873-8248; Web site: http://www.education.gov.uk |
| Peer Reviewed: | N |
| Page Count: | 236 |
| Publication Date: | 2026 |
| Document Type: | Reports - Research |
| Education Level: | Secondary Education |
| Descriptors: | Foreign Countries, Secondary Education, Exit Examinations, Scores, Secondary School Students, Special Needs Students, Special Education, Regular and Special Education Relationship, English Curriculum, Mathematics Achievement, Academic Achievement |
| Geographic Terms: | United Kingdom (England) |
| ISBN: | 978-1-83870-753-8 |
| Abstract: | This report presents findings from an analysis of GCSE outcomes for pupils with Education, Health and Care Plans (EHCPs), comparing those educated in mainstream schools with those in special schools. The analysis covers thirteen academic years from 2011/12 through 2023/24, using data from the National Pupil Database (NPD). Key finding: After controlling for prior attainment, type of Special Educational Need and Disability (SEND), and demographic characteristics, pupils with EHCPs in mainstream schools achieve on average 0.564 grades higher across their combined English and Maths GCSEs than comparable pupils in special schools (95% CI: 0.52-0.607). This comparison is based on the 94,927 pupils with EHCPs who took GCSEs between 2011/12 and 2023/24 -- representing approximately 22% on average of all pupils with EHCPs. Among these GCSE-taking pupils, on average across the study period, nearly all in mainstream (96.7%) and most in special schools (70.2%) were entered for both English and Maths. This finding is consistent across multiple statistical methods, with estimates ranging from 0.54 to 0.60 grades. Importantly, this comparison controls for prior attainment at Key Stage 2 (KS2), meaning we are comparing pupils who demonstrated similar academic ability at age 11. However, this finding requires careful interpretation. The analysis identifies an association, not a causal effect. Pupils are not randomly allocated to school settings; placement decisions reflect individual needs assessments. Special schools serve pupils with the most complex requirements, and our data cannot fully capture severity of need within each SEN category. The observed difference may be partially explained by unmeasured differences in severity of need between pupils in different settings, though sensitivity analyses (E-value = 1.87) suggest the entire effect is unlikely to be solely due to unmeasured confounding. The mainstream association varies by type of need. It is largest for pupils with Social, Emotional and Mental Health (SEMH) needs (0.68 grades), Physical Disabilities (PD; 0.65 grades) and Specific Learning Difficulties (SpLD; 0.63 grades). One possible explanation for the larger SEMH association is that social, emotional and mental health difficulties often emerge or intensify during adolescence, meaning their impact on secondary school outcomes may not be fully captured by prior attainment at Key Stage 2. It is smallest for pupils with Moderate Learning Difficulties (MLD) and Speech, Language and Communication Needs (SLCN; both 0.44 grades). These findings hold after accounting for Local Authority (LA) variation in EHCP rates and placement patterns through multilevel modelling. Local Authorities vary substantially in both EHCP identification rates and placement decisions, so this analysis helps rule out geographic or local policy differences as explanations for the observed outcomes. The relationship has changed over time. The mainstream association widened from 0.561 grades pre-2014 to 0.632 grades following the Children and Families Act reforms, before narrowing to 0.465 grades in the post-COVID period. These temporal patterns warrant further investigation. The post-COVID narrowing is particularly notable. Whether this reflects lasting changes to the relationship between setting and outcomes, or temporary factors including the use of teacher-assessed grades during the pandemic, remains unclear. Important limitations: This analysis covers only approximately 22% on average of pupils with EHCPs who take GCSEs (averaging across the 2011/12-2023/24 period). Findings may not generalise to pupils with the most complex and profound needs. Regarding the "inclusion debate", this analysis examines outcomes only for pupils already identified as requiring substantial support -- it does not speak to the broader question of inclusion for all pupils with additional needs. GCSE outcomes represent only one dimension of educational success; wellbeing, social development, life skills, and vocational qualifications -- which pupils with SEND may be more likely to pursue -- are not captured by this analysis. Additionally, changing placement patterns over time -- with increasing numbers of pupils with EHCPs in both settings -- may affect the composition of each group, potentially explaining some of the temporal variation in estimates. These findings represent averages across a large sample of pupils. They do not provide a basis for individual placement decisions, which must continue to reflect comprehensive assessment of each pupil's needs, circumstances, and the full range of outcomes that matter for their education and development. |
| Abstractor: | As Provided |
| Entry Date: | 2026 |
| Accession Number: | ED679941 |
| Database: | ERIC |
| Abstract: | This report presents findings from an analysis of GCSE outcomes for pupils with Education, Health and Care Plans (EHCPs), comparing those educated in mainstream schools with those in special schools. The analysis covers thirteen academic years from 2011/12 through 2023/24, using data from the National Pupil Database (NPD). Key finding: After controlling for prior attainment, type of Special Educational Need and Disability (SEND), and demographic characteristics, pupils with EHCPs in mainstream schools achieve on average 0.564 grades higher across their combined English and Maths GCSEs than comparable pupils in special schools (95% CI: 0.52-0.607). This comparison is based on the 94,927 pupils with EHCPs who took GCSEs between 2011/12 and 2023/24 -- representing approximately 22% on average of all pupils with EHCPs. Among these GCSE-taking pupils, on average across the study period, nearly all in mainstream (96.7%) and most in special schools (70.2%) were entered for both English and Maths. This finding is consistent across multiple statistical methods, with estimates ranging from 0.54 to 0.60 grades. Importantly, this comparison controls for prior attainment at Key Stage 2 (KS2), meaning we are comparing pupils who demonstrated similar academic ability at age 11. However, this finding requires careful interpretation. The analysis identifies an association, not a causal effect. Pupils are not randomly allocated to school settings; placement decisions reflect individual needs assessments. Special schools serve pupils with the most complex requirements, and our data cannot fully capture severity of need within each SEN category. The observed difference may be partially explained by unmeasured differences in severity of need between pupils in different settings, though sensitivity analyses (E-value = 1.87) suggest the entire effect is unlikely to be solely due to unmeasured confounding. The mainstream association varies by type of need. It is largest for pupils with Social, Emotional and Mental Health (SEMH) needs (0.68 grades), Physical Disabilities (PD; 0.65 grades) and Specific Learning Difficulties (SpLD; 0.63 grades). One possible explanation for the larger SEMH association is that social, emotional and mental health difficulties often emerge or intensify during adolescence, meaning their impact on secondary school outcomes may not be fully captured by prior attainment at Key Stage 2. It is smallest for pupils with Moderate Learning Difficulties (MLD) and Speech, Language and Communication Needs (SLCN; both 0.44 grades). These findings hold after accounting for Local Authority (LA) variation in EHCP rates and placement patterns through multilevel modelling. Local Authorities vary substantially in both EHCP identification rates and placement decisions, so this analysis helps rule out geographic or local policy differences as explanations for the observed outcomes. The relationship has changed over time. The mainstream association widened from 0.561 grades pre-2014 to 0.632 grades following the Children and Families Act reforms, before narrowing to 0.465 grades in the post-COVID period. These temporal patterns warrant further investigation. The post-COVID narrowing is particularly notable. Whether this reflects lasting changes to the relationship between setting and outcomes, or temporary factors including the use of teacher-assessed grades during the pandemic, remains unclear. Important limitations: This analysis covers only approximately 22% on average of pupils with EHCPs who take GCSEs (averaging across the 2011/12-2023/24 period). Findings may not generalise to pupils with the most complex and profound needs. Regarding the "inclusion debate", this analysis examines outcomes only for pupils already identified as requiring substantial support -- it does not speak to the broader question of inclusion for all pupils with additional needs. GCSE outcomes represent only one dimension of educational success; wellbeing, social development, life skills, and vocational qualifications -- which pupils with SEND may be more likely to pursue -- are not captured by this analysis. Additionally, changing placement patterns over time -- with increasing numbers of pupils with EHCPs in both settings -- may affect the composition of each group, potentially explaining some of the temporal variation in estimates. These findings represent averages across a large sample of pupils. They do not provide a basis for individual placement decisions, which must continue to reflect comprehensive assessment of each pupil's needs, circumstances, and the full range of outcomes that matter for their education and development. |
|---|---|
| ISBN: | 978-1-83870-753-8 |
