Profiling SEN and Disability in NI Using Education and Social Data

This research used educational and social data to explore Special Educational Needs in NI and the relationship between SEN and disability relative to wider socio-economic influences.


Overview

This project (Project No: ES/S00601/X1) was funded by the Economic and Social Research Council, Secondary Data Analysis Initiative (SDAI), conducted under the auspices of the UK’s Administration Data Research Partnership, and completed by a multi-disciplinary team comprising researchers from Ulster University and the University of Leicester.

The SDAI gives accredited researchers access to linked, de-identified cross-departmental administrative data in a secure environment, thereby supporting research to improve knowledge, policymaking and public service delivery.

The numbers of pupils with SEN in NI have increased in the past ten years at a rate that is proportionately higher than in the general school population; at the same time, the association between SEN prevalence and wider environmental factors has been an enduring concern in NI, particularly where co-existing deprivation indicators – such as poverty, housing conditions and unemployment – are key determinants on outcomes for children.

Project Aims

  • To interrogate the utility of existing education data sources as a means of gaining insights into pupils with SEN in NI.
  • To interrogate the utility of existing social data sources to gain insights into the relationship between SEN and disability relative to wider socio-economic influences.

Method

Data was obtained from two sources: the Department of Education (DE) Research and Statistics Branch, and the Northern Ireland Statistics and Research Agency (NISRA) through the Northern Ireland Longitudinal Study (NILS).

DE data analyses drew on ten data sets covering the period 2010/11 – 2018/19; these provided information on pupils with SEN using a range of disaggregated variables, including: education region, school type (primary, post-primary and special), SEN Stage, over-arching SEN category, individual SEN type, gender and year group.

Additional data relative to pupils’ SEN status (Free School Meal Entitlement, Multiple Deprivation Measure, truncated address by Local Government District) were also provided. Data was analysed to record and compare the prevalence of SEN using the range of variables at NI and regional levels and to illustrate the degree of change in these rates over the eight-year period.

NILS data analyses drew on all male and female members aged between 4-19 years enumerated at the 2011 Census and also returned in the 2011 School Census (as well as members of households enumerated in the 2011 Census and linked with those also returned in the 2001 Census).

Analysis was undertaken using cross-sectional analyses of individual, household deprivation and school level predictor variables with health/disability outcome variables in the full 2011 sample, with mixed effects binary logistic regression models applied to examine the unique associations between predictor and outcome variables.

Longitudinal analyses were also conducted to estimate which health and disability variables in 2011 were predicted by individual- and household-level factors in 2001.

Key Findings

There has been a proportionately higher increase (21%) in the overall numbers of pupils with SEN compared to the wider pupil population.  This increase was higher in post-primary schools (26%) than in primary schools (16%), with substantive growth evident in special schools (34%).

Of the seven over-arching categories used to record SEN, the majority of pupils were located in three of these: Cognitive and Learning; Social, Emotional and Behavioural Difficulties; and Communication and Interaction.

Within the category of Communication and Interaction:

  • Prevalence rates were highest in the Belfast region although there was a greater change ratio in the North Eastern region.
  • Prevalence rates were consistently higher among males across school types, with the exception of females at Stages 1-4 in secondary schools in the Belfast region.
  • Prevalence rates for Dyslexia/SpLD, Moderate Learning Difficulties and Mild Learning Difficulties were highest in secondary schools and change ratios were generally higher among females across school types.

Within the category of Social, Emotional and Behavioural Difficulties:

  • Prevalence rates were highest in the Belfast region although there was a greater change ratio in the Western region.
  • Prevalence rates were consistently higher among males across school types, with the exception of females at Stages 1-4 in secondary schools in the Belfast region.
  • Prevalence rates for Attention Deficit Hyperactivity Disorder (ADHD) and Social, Emotional and Behavioural Difficulties (SEBD) were higher in secondary and special schools and change ratios were generally higher among females across school types.

Within the category of Communication and Interaction:

  • Prevalence rates were highest in the Belfast region, with the greater change ratio in the Belfast and North Eastern regions.
  • Prevalence rates were consistently higher among males across school types, although the highest change ratios were among females in the majority of instances.
  • Prevalence rates for Autism Spectrum Disorder (ASD) were highest in primary and special schools and for Speech and Language Difficulties in primary and secondary schools.  Change ratios for ASD were highest among females across school types.

Cross-sectional analysis showed that:

  • Children deprived in employment, tenure and education, children enrolled at schools with below average attendance and above average rates of Free School Meal Entitlement (FSME); children in single parent households and male children were more likely to have a learning, intellectual, social or behavioural difficulty.
  • Children deprived in employment and living in social housing, children enrolled in schools with below average attendance and above average FSME and male children were more likely to have a communication difficulty.
  • Children deprived in household employment and living in social housing, children from single parent households and male children were more likely to have an emotional, psychological or mental health condition.

Conclusions

The availability of administrative data has been a step-change in social science research, bringing cross-cutting reach to benefit academics from a range of disciplines and reinforcing the value of data sharing across the UK and beyond.

In the short term, findings will provide a useful evidence base upon which to build future research and facilitate future data sharing as new data becomes available and accessible; in the longer term, a key legacy goal will be to encourage and facilitate further data sharing in NI.