What would you like to look for?

What is administrative data?
  1. Home
  2. Our mission

Changes in school performance and involvement in the criminal justice system

Categories: ADR England, Office for National Statistics, Children & young people, Research using linked data, Research findings, Impact, ADR UK Research Fellows

30 April 2025

Changes in school performance and involvement in the criminal justice system

Author: Dr Alice Wickersham

Date: April 2025

Research summary

This project used secure education and justice data to explore changes in school performance and involvement in the criminal justice system. Findings suggest that pupils showing relative declines in their school performance throughout their school career were more likely to be convicted or cautioned for criminal offences during young adulthood.  Changes in school performance as early as primary school could help to identify pupils who are struggling and in need of additional support.

These findings have been shared with the Department for Education (DfE) and Ministry of Justice (MoJ). They informed the creation of a short animated video to raise awareness, understanding, and support for administrative data research.

Data used

This project leveraged the de-identified MoJ-DfE linked dataset – England, an ADR UK flagship dataset.

Sociodemographic, school performance, and social care data was extracted from the National Pupil Database. Information on offence convictions and cautions was extracted from Police National Computer data.

Methods used

This project focused on 4.3 million pupils born between 1 September 1990 and 31 August 1997. We investigated their school performance at three key timepoints when statutory testing takes place in England:

  • Year 2 (key stage 1, typically assessed ages 6-7 years)
  • Year 6 (key stage 2, typically assessed ages 10-11 years)
  • Year 11 (key stage 4, typically assessed ages 15-16 years)

Standardised scores were produced at each timepoint. A positive standardised score above zero indicates above-average performance compared to other pupils in the year, while a negative standardised score below zero indicates below-average performance compared to other pupils in the year.

To understand how pupils’ school performance changed over time, we used growth mixture modelling. The resulting trajectories were then used in multilevel logistic regression models to understand their association with criminal justice outcomes, taking into account sociodemographic characteristics and clustering within schools.

Title reads: Five school performance trajectories derived from Growth Mixture Modelling (N=4,317,436) The figure shows the shape of the five trajectories on a line graph, as described in the bullet points. A dashed horizontal line is also provided as a reference line. Labels read: • Average Consistent, N=3,497,167, 81.0% • Average Increasing, N=66,383, 1.5% • Average Declining, N=373,117, 8.6% • Low Increasing, N=98,805, 2.3% • Low Consistent, N=281,964, 6.5%

Research findings

Pupils could generally be described as following one of five possible school performance trajectories as shown in the graph.

Two of these groups particularly stood out:

  • The “Average Declining” group, where pupils started with average performance but declined over time
  • The “Low Consistent” group, where pupils consistently performed below average.

These two groups were the most likely to receive cautions or convictions. For example, in the Average Declining group:

  • 1 in 3 pupils received a caution or conviction before finishing Year 11
  • 1 in 10 pupils received a first caution or conviction as young adults.

The multilevel logistic regression models, which took sociodemographic characteristics and school-level clustering into account, also confirmed that pupils in the Average Declining and Low Consistent groups were at increased risk of first convictions or cautions as young adults, as compared to pupils in the Average Consistent group.

Research impact

The findings from this project have been presented widely, including to the DfE and MoJ. They were also incorporated into an animated video about administrative data research for adolescents and young adults. Education data is being collected for pupils and students now, so it is essential to raise awareness, understanding, and support for administrative data research in this age group.

Investigations from the wider project, such as the quality and consistency of ethnicity data in the linked dataset, also continue to have an impact. They were cited by the Children’s Commissioner and led to the production of a learning resource for ADR UK.

This work feeds into a wider programme of research which also considers the role of child and adolescent mental health, led by the CAMHS Digital Lab. It has led to further funding to extend the work, including an ADR UK PhD Studentship to investigate education and offending trajectories among female offenders.

Research outputs

Publications, reports and guides

Blogs, news posts, and videos

Presentations and awards

Acknowledgements

With thanks to co-investigators Dr Rosie Cornish, Dr Johnny Downs, and Professor Stephen Scott.

About the ONS Secure Research Service

The ONS Secure Research Service is an accredited trusted research environment, using the Five Safes Framework to provide secure access to de-identified, unpublished data.

If you use ONS Secure Research Service data and would like to discuss writing a future case study with us, please get in touch at IDS.Impact@ons.gov.uk. Please also report any outputs here: Outputs Reporting Form

 

Documents

Share this:

Cookie Extension Logo