Trajectories of female violent offending

Why we funded this

Research has indicated that there are two main offending trajectories: life-course persistent, which starts at a young age and persists into adulthood, and adolescent-limited, which starts in the mid-teens but quickly stops. Most previous work on violence has used either male-only samples or combined samples in which males dominate. It is not clear whether females, particularly those who are involved in violent offending, follow different trajectories. 

Individuals from ethnic minority backgrounds, those with care experience, special education needs, and those from deprived backgrounds are over-represented in the criminal justice system. However, there is some evidence to suggest that the extent of increased risk in these subgroups may differ between males and females.  

This project will inform efforts to identify girls and young women at risk of violent offending and will inform the design and targeting of interventions aims at preventing girls and young women getting involved in violence.  

What are the main research questions

The main research questions are: 

• What are the offending trajectories between the ages of 10 and 24 years among women and girls in England convicted or cautioned for violent offences and how do these compare to trajectories seen in males? 
• Do these trajectories differ between ethnic groups, and across vulnerable subgroups such as those with care experience, educational needs and those experiencing deprivation? Are these relationships similar in females and males? 
• What are the trajectories into violent offending, key transition points, and the potential impact of interventions on girls’ behaviour from the perspectives of girls/young women, youth justice and probation professionals? 

What the analysis involves 

This project will use linked data from the National Pupil Database and the Police National Computer. This includes school data on all children in state-funded schools in England, and records on convictions and cautions.  

Longitudinal latent class analysis will be used to classify individuals into groups according to their offending patterns over time (trajectories). Multinomial logistic regression will be used to examine whether different ethnic groups and those with various vulnerabilities are more or less likely to follow these trajectories 

The project focuses mainly on females who have been convicted/cautioned for at least one violent offence, but comparisons with males will be made. 

Interviews will be conducted with 10 young women who have been cautioned/convicted for violent offending, and online focus groups will be held with relevant professionals. This qualitative data will be analysed thematically to illuminate key transition points and interventions that might affect offending trajectories.