Government needs to generate better evidence on SEND provision as part of the reforms
The government is about to announce major changes to the SEND service, but how strong is the evidence for the principles guiding the reforms?
The HOPE study (Health Outcomes for young People throughout Education), published in a report by researchers at UCL and Cambridge University, provides qualitative evidence that the SEND services can work well. Reported experiences by young people and parent-carers of SEND provision suggest that timely, good quality, tailored SEND provision can improve engagement with lessons and with school, anxiety management, and children feel happier, less distressed, and better able to socialise.
While these findings tell us what can work, they are not the same as quantitative evidence showing that the SEND service as delivered in practice benefits children overall. Such evidence is limited and requires specific questions being tested in trials or using statistical approaches to emulate trials.
The HOPE study used the ECHILD database for all children in England to emulate trials using past health and school records from children more than a decade ago and followed up since then. The study showed that for certain groups, such as children with cerebral palsy, SEND provision in Year 1 of primary school reduced unauthorised absences but did not improve school test scores or reduce hospital use.
The researchers emphasise that these methods underestimate the benefit of SEND provision. This is partly because severity of need affects who receives SEND provision but is not measured in the ECHILD data. Another reason is that no information is captured on what support is received. This can range from extra time for tasks through to one-to-one support from a teaching assistant.
Government could generate stronger evidence by implementing changes to SEND services in a systematic way. For example, schools could be randomised to early provision in nursery compared with usual practice, or to intensive training of teachers and teaching assistants compared with usual training. Stepped implementation or other randomised designs, combined with improved data, would provide stronger evidence. This way government would generate evidence on what works and what does not so should be changed as reforms progress.
The HOPE study demonstrated how ECHILD data could be used by government to monitor whether who gets SEND provision is fair and according to need. Nearly 1 in 3 children had SEND provision by age 11 and this was mostly driven by differences in pupil’s health conditions, social disadvantage, school test scores and type of school. Voluntary schools and academy sponsor-led schools were less likely than community schools to provide the more intensive form of SEND provision, called an education, health and care plan (EHCP). There was very little variation in SEND provision between local authorities, after adjusting for these other factors.
Contact Professor Ruth Gilbert [email protected] or Professor Tamsin Ford [email protected]
For more information on the HOPE study click here and to view all published articles click here