New dataset improves linkage of some of the most vulnerable children in Wales
A team of researchers from Swansea University have developed a novel algorithm that enhances the availability to link administrative data from some of the most vulnerable children in Wales: those who are looked after (children in care). The algorithm makes use of several readily linkable administrative datasets accessible within the Secure Anonymised Information Linkage (SAIL) Databank, to boost the number of unique identifiers (variables used to link datasets) that are allocated to a vulnerable child or young person.
This algorithm has been used to produce a powerful new dataset that has recently been made available to researchers. It promises to advance our knowledge about children looked after by enabling linkage of data from cross-sector agencies such as health, education and criminal justice. Administrative data from various sectors has the potential to provide a more coherent picture of how children navigate through different services.
Researchers wishing to link Welsh, English and Scottish data from children aged four and under who are involved in children’s statutory social services face a major obstacle. Currently, the identifiable information collected about children looked after by the local authority is limited. Instead, additional identifiable information must be obtained from the education dataset to allow for a child to be linkable. This is a major shortfall as it has been shown that children under the age of five are less likely to re-enter care compared to older children. This means that children who do not re-enter care after they reach school age will never be able to have their data linked.
A new open access article published in the International Journal of Population Data Science (IJPDS), provides an overview of this work. It describes how the research team were able to improve the number of children and young people who were linkable across different datasets from 36.6% to 61.2%. There was some evidence to show that demographics such as ethnicity may impact whether a child or young person can be linked. By improving the number of linkable children and young people, this work was also able to reduce some of the bias that occurs with missing data.
Grace Bailey, lead author from Swansea University added, “The availability of additional identifiers provides huge potential for furthering our understanding of children looked after, particularly among children under 5 years old. While the algorithm was specifically designed with Welsh data at the forefront, we hope that researchers from other countries will go on to adapt it for their own data and research. We believe that this new dataset will open up new opportunities for researchers to generate evidence and inform policy and practice.”
Dr Grace Bailey, Research Officer, Swansea University
Bailey, G., Lee, A., Ahmed, S., Cowley, L., Stuart, A., Farr, I., Brooks, C., North, L. and Griffiths, L. (2025) “Improving opportunities for data linkage within Children Looked After administrative records in Wales”, International Journal of Population Data Science, 10(1). Available at: https://ijpds.org/article/view/2383 (Accessed: 19 February 2025).