We explored the presence of relatively discrete subgroups of adolescent females in a statewide sample (N = 94) using latent class statistical methodology. Specifically, we hypothesized that, rather than being a homogenous group, the girls in the sample would separate into clusters with both crossover and distinguishing characteristics. Results showed confirmation of our hypothesis with a best-fit three cluster solution consisting of a low violence cluster, a moderately violent cluster, and a high-rate violent cluster of girls. Findings provide a more precise identification and description of subgroups of female juvenile offenders to inform treatment efforts and provide a preliminary typology/cluster profile to build upon in future research addressing this population.
All Science Journal Classification (ASJC) codes
- Pediatrics, Perinatology, and Child Health