Using routinely-collected educational data to identify health needs and vulnerabilities
The UK government has used the word crisis to describe the current processes for identifying and supporting children with educational needs (e.g., autistic children). In the NIHR ARC Yorkshire and Humber, researchers at the Centre for Applied Education Research (CAER), the University of Leeds and Lancaster University are exploring how linked routine data can be harnessed to support earlier identification of potential needs.
In England, children are assessed at school entry through the Early Year Foundation Stage Profile (EYFSP). This includes academic and non-academic (e.g., motor and social skills) measures of school readiness. The non-academic elements share a lot of similarities with the items used to assess children for developmental conditions. This suggests that a routine, statutory assessment conducted on all children when they are 4-5 years of age might be a very useful tool for identifying health needs and vulnerability.
Indeed, previous research from our group has shown that the EYFSP can help identify children with Special Educational Needs (SEN). Children who did not reach a ‘good level of development’ on the EYFSP were much more likely to be identified as having SEN than children who did reach a good level of development. A new study, led by Dr Megan Wood, Dr Amy Atkinson, Dr Lydia Gunning, and Prof Mark Mon-Williams builds on this research using data from over 50,000 children between 4 and 12 years of age. These data were obtained from Connected Bradford, a large population database.
The first aim was to further examine whether the EYFSP can identify children who are at increased risk of being identified as having SEN. We found clear evidence of this. Strikingly, almost half of children (49%) who did not reach a good level of development were later identified as having SEN. In contrast, less than 10% of children who reached a good level of development required additional support. Children who did not reach a ‘good level of development’ were therefore over 5x as likely to require SEN support than children who reached a good level of development.
The second aim was to explore whether the extent to which the EYFSP can identify children with SEN varies depending on the child’s ethnicity. We found evidence to support this.
Importantly, in children who did not reach a good level of development, identification of SEN was over 16% higher in White British children, compared to children from ethnic minority backgrounds (53.3% of White British children identified as SEN vs. 45.6% of children from ethnic minority backgrounds).
Further analyses then examined whether these differences were a reflection of inequalities in the timeliness of SEN identification. If this was the case, we might expect White British children who did not reach a good level of development to be more likely to be identified with SEN earlier in development, but for these differences to later disappear. Indeed, this is what we observed. At Key Stage 1 (5-6 years), identification of SEN in children who did not reach a good level of development was over 16% higher in White British children compared to children from ethnic minority backgrounds (53.4% of White British children identified as SEN vs 46.0% of children from ethnic minority backgrounds). No significant differences were observed at Key Stage 2 or Key Stage 3 after other demographic information (such as socioeconomic status) were taken into account.
These findings suggest that White British children at higher risk of SEN are more likely to be identified earlier than children from ethnic minority backgrounds. This may be because of an implicit ethnic bias in the identification of children with SEN, cultural biases in family perceptions, or families lacking the social capital to push for additional support. These findings build on previous research that show delays in the identification of a range of conditions for ethnic minority groups.
This research highlights the immense opportunity of using the EYFSP to identify children who are more likely to require additional support from an early age. This may be especially useful for children from ethnic minority backgrounds, who often experience significant delays in the identification of SEN.
Based on these findings, we are now developing a purpose-built Electronic Developmental Profiling Tool (EDPT) in collaboration with educational psychologists. This teacher-reported assessment will evaluate both academic and non-academic abilities in children and young people across the school years. This is based on the EYFSP and other developmental screening tools. The tool has already sparked ministerial interest and will be implemented across the Bradford District and beyond to allow teachers to identify additional needs across development. This will ensure targeted and timely support can be put into place. For updates on the development of this tool and similar work, please see the CAER website (https://caer.org.uk/).
We argue that using routinely collected educational data such as the EYFSP can empower schools and teachers with the insight they need to ensure every child can realise their full potential.
This work is currently under review at an academic journal, and we expect that it will be published soon. For updates about this, please see the CAER website. Our previously published paper demonstrating a relationship between the EYFSP and SEN can be found here: https://www.sciencedirect.com/science/article/pii/S0959475221000967.
This blog was written by by Dr Megan Wood (University of Leeds), Dr Amy Atkinson (Lancaster University), Dr Lydia Gunning (University of Leeds, Centre for Applied Education Research, Bradford Institute for Health Research), and Mallory Morehead (Centre for Applied Education Research), part of the Healthy Childhood theme at Yorkshire & Humber ARC.
19 June 2023