Nearly half of people born in Wales in 2002-2003 were classed as having special educational needs (SEN), our new research has indicated, raising questions about the system used to diagnose a generation of Welsh children.

Our report for the Nuffield Foundation found that 48% of this group, who are now aged 20 to 22, were identified as having SEN at some point before they turned 17. In some cases, this may have negatively affected their educational outcomes.

Pandemic disruptions meant complete data was only available for this year group. However, we also identified several factors that made some children born in Wales between 2002 and 2008 more likely to receive an SEN diagnosis – including being a boy, being born in summer, and being on free school meals.

Our findings suggest children from these groups may have been over-identified (and those not in these groups potentially under-identified). A new system for identifying educational needs was introduced in Wales in 2020, and the number of children being diagnosed has since fallen significantly – it was 20% lower in the year after the new system began.

Our findings suggest the previous system was not effectively supporting learners with SEN to achieve academically. We found the more time a child spent with an SEN diagnosis during their education, the less likely they were to to meet nationally expected levels of attainment.

SEN are identified when a child has much greater difficulty learning than most of their peers, or a disability that limits their use of typical educational facilities. This may include autism, dyslexia, attention deficit hyperactivity disorder (ADHD), and moderate learning difficulties.

Our research, funded and reviewed by the Nuffield Foundation and scrutinised by our research advisory group (including academics, policy-makers and practitioners), analysed education data from over 200,000 children in Wales born between 2002 and 2008. We wanted to understand who was being identified as having special educational needs, and the impact this had on their academic achievement.

We found some key common factors that increase the likelihood of children being identified as having additional learning needs for all the children in the study, born between 2002 and 2008.

Boys were 5.5 times more likely to be identified with SEN than girls. While it’s possible boys may be more likely to have learning needs, there may be gender bias at play too, meaning boys’ behaviour gets more attention. Girls may also be better at masking their special educational needs.

Poverty also played a large role. We found that children who had free school meals throughout their education were four times more likely to be identified with educational needs, compared with those not receiving free meals. Alongside this, children born in the most deprived neighbourhoods were shown to be even more likely (4.6 times) to be identified with learning needs.

What’s more, learners born in the summer (so younger in their year group) were three times more likely to be identified with SEN than those born in the autumn.

These findings highlight how a child’s environment can contribute to the identification of educational needs. They also raise concerns about the effectiveness of identification processes, particularly given the unexpectedly high number of learners identified with SEN.

It suggests a potential issue of over- or under-identification of certain children. Younger children may not actually be more likely to have additional learning needs – they might just be behind due to their age. It is crucial to understand a child’s environment and their individual situation to effectively support their learning needs.

We also found the earlier that special educational needs were recognised, and hence the longer a child’s education was spent with these known needs, the less likely children were to meet nationally expected levels of attainment. This shows that whatever additional support children identified as having educational needs were receiving, it was not effective in mitigating the impact of their learning needs on achieve their learning goals.

Our research mirrors similar national findings. Evidence from the Education Policy Institute in England also found a high number of children – 39% – in the cohort taking their GCSEs in 2016 had been identified with special educational needs and disability (SEND) at some point in their schooling. Other research has shown that children with SEND in England are far less likely to meet expected learning standards than their peers at Key Stage One.

Recognise diverse learning needs

However, our research also raises questions about the current identification system being used in Wales. While some children in the 2002-03 cohort may have been over-identified, the fact these children were identified as having additional needs means their schools felt they required extra help. These children may go under the radar in the new system.

We believe a more inclusive model for supporting children’s learning is required, which takes their individual circumstances into account when providing help.

An effective system should give support that allows children with additional learning needs to meet the national expectations, at a minimum. To improve academic attainment levels in Wales, it is crucial to prioritise effective support for this potentially very large group of learners.

Inclusive educational initiatives that recognise and support children’s diverse learning needs are necessary. By acknowledging that children can have different learning needs at different times, schools can consider how they can support all learners – not just those who are identified with additional learning needs.

The reforms introduced by the Welsh government are changing the way children are identified and supported. We believe the methods used to identify learning needs should be rigorously reviewed, with a new focus on ensuring accuracy, fairness and inclusivity. Environmental factors such as the child’s age in a year group should be taken into consideration when identifying their learning needs.