Links to preceding blogs in the series
Blog #1: Intersections between ethnicity and additional needs
Blog #2: Additional provision for SEND, ethnicity and attainment
Blog #3: Staffing variations for EAL, ethnicity and attainment
Recap of the aims and findings
In this blog series we have considered the relative GCSE English and maths attainment of pupils with additional needs during the Covid-19 pandemic – those with Special Educational Needs and Disabilities (SEND) and those who speak English as an Additional Language (EAL) – and in particular, we focused on comparisons between pupils with similar additional needs but differing ethnic backgrounds.
The context for the analysis presented in this blog series was the first year of the Covid-19 pandemic which resulted in the use of GCSE grades based on centre-assessed grades, i.e. grades that were decided by teachers and moderated by school leaders, since examinations were cancelled.
Our focus on additional needs was motivated by the increased GCSE attainment gaps experienced by pupils with EAL or SEND in 2020, and our analysis by ethnicity has revealed further variation in outcomes within these groups.
In relation to pupils with SEND, we found that:
▪ In 2020, pupils with 6+ years of School Support had similarly low GCSE results to pupils with Education Health and Care Plans (EHCPs) who receive greater support for their needs and the protection of a statutory plan.
▪ This suggests that this ‘shadow EHCP’ group of pupils with 6+ years of School Support are likely to need better support, particularly at and after the transition to secondary school.
▪ Across ethnic groups, attainment was much more similar for pupils with SEND than it was for all pupils in each ethnic group, especially among pupils with EHCPs or 6+ years of School Support. This suggests that SEND needs account for important variation between ethnic groups.
In relation to pupils who speak EAL, we found that:
▪ In 2020, most ethnic groups among pupils who speak EAL and arrived during secondary school (years 7 to 11), had attainment below the national median in GCSE English and maths.
▪ However, attainment for pupils who speak EAL is nuanced; those who arrived in school in England before Year 7 had good attainment, but in contrast for the three lowest-attaining ethnic groups (Gypsy Roma, Traveller Irish and Black Caribbean) and White & Caribbean pupils, late arrival in Years 10-11 resulted in attainment in the bottom quarter of the national distribution.
▪ There was more variation in attainment between ethnic groups for pupils who speak EAL than for pupils with SEND. This is likely to reflect differences in educational and life experiences prior to arriving in England, as well as important differences in English language proficiency on arrival.
▪ This variation meant that the largest attainment gaps for pupils who speak EAL were similar in size to the largest SEND attainment gaps. Yet there is almost no policy focus on EAL in England.
After exploring the GCSE attainment outcomes of different intersectional subgroups of pupils with varying recognised additional needs across ethnic groups, we extended our analysis to show how attainment differed or was similar for pupils of different ethnicities whose schools made additional provision for additional needs or had different staffing, and those whose schools did not.
The types of provision for SEND we considered are whether the school had a SEN Resourced Provision, or a SEN Unit; and whether or not the school was among those with the most teaching assistants (in the top quartile for the ratio of teaching assistants to teachers). We found that:
▪ In 2020, effects of additional provision were largest for pupils with EHCPs (with the greatest needs) but remained larger for pupils with 6+ years of School Support than 1-5 years of support.
▪ Resourced Provisions were most associated with better attainment of pupils with 1-5 years of School Support, while both SEN Units and Resourced Provisions were equally associated with positive effects for pupils with 6+ years of School Support.
▪ For pupils with EHCPs, SEN Units, but mainly higher ratios of teaching assistants, were associated with better attainment.
▪ However, benefits of additional provision were not uniformly spread across ethnicities, nor were they always present for low-attaining ethnic groups. Some of the gaps in positive effects may be due to small sample sizes, but this would not account for all the differences observed.
▪ For example, it’s not clear why Black Caribbean pupils with EHCPs benefited from attending a school with a SEN Unit (+28 GCSE percentiles, net) to a greater extent than Black African pupils (+11 GCSE percentiles).
▪ Nor is it clear why Other Asian pupils with 6+ years of School Support benefited from attending a school with a high quotient of teaching assistants (+4 percentiles) but comparable Indian pupils did not appear to benefit. These comparisons of ethnicities are within the same broad ethnic categories and sit near one another in the distribution of attainment of pupils with SEND.
▪ Understanding these differences is important for the successful and equitable delivery of any reforms to the SEND system, whether those of the current SEND and AP Improvement Plan, or those introduced by any future government.
▪ It is not possible to answer every question of interest through data analysis alone. Future research on SEND and ethnicity would benefit from the inclusion of qualitative methods to unpick what is happening on the ground in greater detail.
For EAL, we considered different staffing profiles of the school attended by pupils who speak EAL: whether the school was in the top quartile with the most teachers of Black and Minority Ethnicity backgrounds, alongside the teaching assistants measure previously described. We found that:
▪ There is no additional provision for pupils who speak EAL mandated by national policy and the funding schools receive in respect of this group is of short duration, lasting only the first three years after arrival, in contrast with the 5-8 years that it takes to develop academic proficiency.
▪ We considered the attainment of pupils in schools with different staffing profiles that might have some positive role modelling benefits for these pupils. This is not comparable with provision under the SEND code of practice but is the nearest measurable factor.
▪ It’s important to note that a school having more BME teachers or more teaching assistants does not mean these staff have specialist training in teaching pupils who speak EAL.
▪ Very few (three) positive effects of different staffing were found for pupils who speak EAL in 2020. These were restricted to pupils who arrived in their first state school in England before Year 10 (i.e. not those at greatest risk of poor GCSE outcomes), and to schools with a high quotient of BME teachers (and not those with a high quotient of Teaching Assistants).
▪ The positive staffing effects may have occurred due to chance, but the largest effect, for Black Caribbean pupils who arrived in Years 7-9 (+19 GCSE percentiles), could be due to ‘ethnicity matching’ between pupils and teachers, or to particular needs of this ethnic group.
▪ There were a much larger number of negative staffing effects (fifteen) spread across all EAL ‘arrival time’ categories, and across schools with high quotients of Teaching Assistants, as well as those with high quotients of BME teachers.
▪ While extremely concerning, these results are not surprising, since compulsory initial teacher training and early career training do not address the pedagogies needed to improve pupils’ English language proficiency as swiftly as possible so that they can benefit from the curriculum.
▪ It’s also possible that these negative effects reflected the unequal impact of the pandemic in 2020, and that pupils who speak EAL got lost in the many learning, mental health and safeguarding priorities that staff in schools have been overwhelmed with.
▪ A further possibility is that teachers who knew that the usual provision was impacted by the pandemic school closures and restrictions for this cohort may have therefore expected the pupils with additional needs to have lower attainment as a result, with the grades awarded confirming these expectations, since grades were not independent of teachers’ beliefs in 2020.
Policy implications for SEND
While this cohort has since completed further education, disadvantages experienced during the pandemic are likely to have implications for their destinations after key stage 5, whether those are in higher education, apprenticeships or employment, or not in education or training.
It’s also important not to forget those young people in the age cohort that we have analysed who will not have appeared in the 2020 GCSE results due to the severity of their SEND needs, many of whom are still in education under the extension of the SEND Code of Practice to age 25.
Education, health and care services were maintained as best providers could arrange during the pandemic, but education providers were often left unsupported by health, social care and local authority services, and in some locations schools did not or could not accommodate children either.
While we are unable to analyse outcomes for pupils without GCSE entries, this underlines the importance of further research into broader outcomes for children and young people with SEND. The findings we were able to provide in relation to the GCSE cohort suggested the following implications for SEND policy:
▪ The existence of the shadow EHCP group (pupils with School Support for six or more years whose attainment was similar to that of pupils with EHCPs) raises new doubts about the sufficiency of current arrangements for SEND provision.
▪ It is clear that for the 2020 GCSE cohort, greater support was needed for pupils with longstanding School Support status, across ethnic groups.
▪ A public commitment is needed not to artificially limit or set targets to reduce the number or proportion of pupils who receive the statutory protection of an EHCP. The consequence of limiting EHCPs is likely to be intensified rationing of support for the most economically disadvantaged pupils in local authorities with greatest need (Hutchinson, 2021).
▪ Our findings call into question the current practice whereby many pupils with School Support during primary school are removed from the SEND register and lose their support when they transition to secondary school.
▪ Since SEND is associated with neuro-developmental differences that are visible in brain scans (e.g. Mareva et al., 2023 preprint), long-standing forms of SEND need are unlikely to simply disappear following the provision of support in primary school. The legacy of this was visible in the 2020 GCSE results of pupils with School Support for six or more years.
▪ The consequences of SEND needs that are not met are not restricted to severely reduced GCSE grades; many pupils find mainstream school difficult to attend, resulting in increased absences from school or in extreme cases total withdrawal from school, including that of children who are looked after by local authorities, often combined with mental health difficulties.
▪ An absence of compulsory training for teachers in SEND and neurodevelopmental differences (the only government-funded training is for autism and is not compulsory) places pupils at a disadvantage in mainstream schools.
▪ We were able to formulate general conclusions about forms of provision that were more beneficial for different categories of pupils with SEND based on patterns of attainment across ethnic groups, but we were not able to explain why these benefits of additional provision for SEND in secondary school were unevenly experienced across ethnic groups.
▪ Further research including the use of mixed or qualitative methods is needed to understand why additional provision is not experienced equally by pupils with SEND across all ethnic groups.
▪ Additional and broader outcomes should be considered in future research in order to include those with the most complex or greatest SEND needs.
Policy implications for EAL
The main implication of our findings for pupils who speak EAL is that there is a pressing need for a coherent policy to address their needs to be updated and re-introduced in England, and for school accountability and funding to be aligned with the needs of pupils who speak EAL.
Since the ‘mainstreaming’ of the Ethnic Minority Achievement Grant in 2011, there has been no policy for these pupils beyond a school funding formula factor that does not meet the developmental needs of pupils as it is not provided for the full length of time that it takes to become proficient in English.
Against this backdrop , the following policy implications have emerged from our findings in relation to the 2020 GCSE cohort:
▪ Deficits in GCSE grades for pupils who speak EAL, arrived in years 10-11 and belonged to low-attaining ethnic groups of a similarly unacceptable magnitude to the largest attainment gaps for pupils with statutory EHCPs for SEND.
▪ Late arrival in schools in England combined with EAL and with the effects of a global pandemic were a perfect storm in 2020, but the problem is not new. The absence of any statutory framework of support for pupils who speak EAL meant that they were not prioritised in general, nor in pandemic mitigations.
▪ The low attainment of late-arriving pupils who speak EAL in 2020, and in other years, is a story about the ‘policy neglect’ of this group, that could not be overcome by schools, even as they made herculean efforts to support as many pupils as possible as much as they could.
▪ Although the clock cannot be turned back for the 2020 GCSE cohort who did not experience the ‘normal’ path to GCSEs, experiencing interruption of their schooling both by migration and by a global pandemic, better funding is needed for future cohorts as the school system continues to navigate many pre-, during- and post-pandemic challenges.
▪ The answer to these problems, setting aside the need for better pandemic resilience in schools, is a national policy for meeting EAL needs in schools, involving:
▪ More intensive financial support for schools and colleges to create bespoke qualification pathways focusing on valuable skills and qualifications by age 19;
▪ Adaptation of the expected milestone of GCSE entries at age 16 for pupils who speak EAL and arrive close to this age, combining core subject GCSEs with carefully-chosen post-16 qualifications over the remaining time to maximise the credentials young people leave education with.
▪ Extra time in funded further education could be considered for the oldest arrivals similarly to age 0-25 provision for SEND;
▪ The creation funded mandatory training, and specialist staff roles to bring a parity of professional provision to pupils whose additional needs are linguistic, embedded through a statutory code of practice;
▪ Reintroduction of mandatory assessment and national data collection of English language proficiency for pupils who speak EAL;
▪ A public commitment that education data, including the English proficiency assessments, will only be used for strictly educational and research purposes and not accessed for immigration control purposes.
▪ A public commitment to regularly publish pupil attainment data with meaningful breakdowns of pupils who speak EAL, including by English proficiency, by time of arrival, by first language and ethnicity.
