10 May 2022
In this blog, ADR UK Research Fellow Alice Wickersham describes her work on the Ministry of Justice and Department for Education linked dataset, exploring the relationships between changes in educational attainment and criminal offending.
The Ministry of Justice (MoJ) and Department for Education (DfE) recently created a large, linked dataset containing de-identified records from the National Pupil Database (NPD) and Police National Computer (PNC). Using this dataset, we can analyse relationships between education and offending behaviour which leads to involvement in the criminal justice system.
With recent funding from ADR UK, I am undertaking a 12-month project using this innovative dataset. I will explore two key areas:
- The relationship between changes in educational attainment throughout school and offending
- The uses and limitations of the linked data.
The importance of changes in educational attainment
Identifying factors that either increase or decrease risk of offending is a key research priority for the Ministry of Justice. Understanding these factors could inform policy initiatives to support young people at risk of engaging in offending behaviour.
Offending is thought to be associated with lower educational attainment. However, very little research has been conducted to understand the role of changes in attainment throughout the school career.
Some of the questions I will aim to answer include:
- Do improving, declining, or consistently high or low grades through school increase risk of offending?
- Does the answer to this question change depending on the type of offence?
- Can grades through school increase or decrease the risk of offending among known vulnerable groups, like looked after children and pupils with special educational needs?
As a result of this research, we may be able to identify children and adolescents who are at increased risk of later offending, and provide appropriate support.
Understanding the uses and limitations of the linked data
While large, linked administrative datasets allow researchers to explore complex relationships between different areas like education and the criminal justice system, using these datasets also presents some challenges.
One challenge is that sociodemographic information is not always consistently recorded in administrative datasets. This can make it difficult for researchers to know which of multiple sociodemographic variables to rely on for their research.
Gender and ethnicity are recorded in multiple variables in the NPD and PNC. I will investigate how completely and consistently these two characteristics are recorded between these datasets. As a result, we may be able to make recommendations for reconciling any discrepancies. These recommendations could inform improvements in how gender and ethnicity are recorded in national datasets.
Another challenge is using administrative datasets to evaluate the effectiveness of treatments and interventions. For example, can we investigate whether receiving special educational needs provision lowers risk of offending? To do this, we would ideally need to compare rates of offending between two groups of individuals: those who received special educational needs provision, versus those didn’t despite having the potential to benefit from it.
We already know from the NPD which pupils received special educational needs provision. But how can we tell which pupils had special education needs which were unrecognised, and therefore not provided for? I will investigate this, building on ongoing work exploring how to create matched control groups in this dataset. As a result, we may be able to conduct a study evaluating whether special educational needs provision is effective for lowering offending risk.
Alice Wickersham is an ADR UK-funded Research Fellow using linked MoJ-DfE data made available through Data First.