The longitudinal association between school performance trajectories and offending behaviour

Understanding risk and protective factors for offending and re-offending is a key research priority area for the Ministry of Justice. A thorough understanding of the circumstances driving offending behaviour is imperative for shaping public policy to reduce criminal activity, which will in turn have societal and economic benefits.

Developed in response to this priority area, this research project will use linked National Pupil Database (NPD) and Police National Computer (PNC) data to investigate school performance as a key predictor of offending behaviour. Low educational attainment is thought to be associated with increased risk of criminal behaviour. But little is known about changes in school performance throughout childhood and adolescence and whether consistently low or declining performance trajectories through school increase offending and re-offending risk. It is also unknown whether consistently high or improving performance through school might mitigate risk for offending among known vulnerable groups, such as those from deprived backgrounds, looked after children, and pupils with special educational needs.

In the proposed research project, I aim to identify subgroups of pupils who follow different school performance trajectories, and investigate their association with different types of offending and re-offending behaviours (such as violence, sexual offences, theft and robbery). I also aim to investigate whether high or increasing school performance trajectories mitigate risk for offending behaviour in vulnerable and marginalised groups, and whether there are regional differences in these relationships. I will achieve this using a longitudinal cohort of pupils born between 1990 and 1997, and using statistical modelling techniques that I have previously developed in NPD data.

Understanding this could inform early identification and prevention strategies for children showing early signs of following certain school performance trajectories, and demonstrate the importance of engaging vulnerable groups in education.

This project will also explore the linked NPD-PNC data resource and make recommendations to future users of the data. In particular, previous data linkages have shown that socio-demographic information is not always consistently recorded between administrative datasets. Therefore, in the course of this project, the NPD and PNC will be checked for discrepancies in gender and ethnicity recordings, and recommendations will be made for reconciling any discrepancies identified. The potential for the data resource to be used to conduct quasi-experimental trials, for example investigating whether special educational needs provision mitigates offending risk, will also be explored. This may lay the groundwork for subsequent advanced fellowship work.

Collaborations and outputs have been designed to maximise the impact of this project. I will consult with professional stakeholders, the Data First User Representation Panel and a local advisory group of young people at various points of project design and analysis. I will also host a stakeholder engagement event to build cross-sector collaborations, help interpret findings, and inform policy implications. Within the funding period, I will focus on producing user-friendly outputs and sharing any code I develop. I will also engage in networking, knowledge exchange, and relevant leadership and engagement training opportunities throughout the funding period.

I will be embedded in a team which has successfully linked NHS mental health records to the NPD, and is currently linking NHS mental health records to the PNC. I will therefore be positioned to use various data linkages for impactful research and, long-term, can extend the ADR UK Data First remit to include three-way data linkages between mental health, education and criminal justice records.