This paper describes the development and validation of a taxonomic classification to unravel the heterogeneity of the female offender population in California Prisons in terms of criminogenic needs and gender responsive factors. The approach uses a comprehensive classification domain that includes gender-sensitive factors (parenting issues, sexual abuse, trauma, poverty, etc) as well as a battery of gender-neutral criminogenic factors from the Northpointe COMPAS Reentry Assessment. Pattern seeking methods were applied to the risk-needs profiles and criminal histories of 718 female offenders. A hierarchical taxonomy with several fairly distinct pathways of female offenders was developed. Three validation approaches were applied: convergence across different quantitative pattern seeking methods, convergence across different samples - using a bootstrapping procedure; and external validation against selected criminal behavior factors. The resulting female offender profile types will be described and their potential implications for both theory and differential treatment briefly noted.

Taxonomic structure is examined in two large samples of delinquent youth in a domain of socio-psychological and personality factors. This paper offers a partial empirical test of the overlapping theoretical taxonomies of Moffitt (Psychol Rev 100:674-701, 1993), Lykken (The antisocial personalities, 1995) and Mealey (1995). The first sample consisted of juvenile offenders (n = 1,572) from three state systems. Multiple cluster analysis methods were applied (Wards method, standard K-means, bootstrapped K-means and a semi-supervised pattern recognition technique). Core or exemplar cases were identified by means of a voting procedure. Seven clusters recurrently emerged across replications. While clear analogues of Moffitt's two main categories were found, several additional stable subtypes emerged that were clearly reminiscent of Lykken's sociopathic, neurotic-internalizing and "normal" types. However, boundaries between types were fuzzy and unstable, and many unclassified cases existed. Internal validation was assessed by cross-method verification. External validation assessed type differentiation on several delinquent behaviors. Finally, generalizability was assessed by repeating the clustering on a large replication sample (n = 1,453) from another state. Six of the seven initial types re-emerged.

Background: Several lines of research have recently examined the possibility of distinct etiological types among delinquents. Moffitt's (1993) developmental taxonomy motivated a number of empirical studies to verify her adolescent limited (Al) and life course persistent (LCP) categories - mostly focusing on boys. However, recent taxonomic studies of causal patterns among female delinquents as well as several criminal career studies, while confirming elements of Moffitt's typology also raise serious challenges - in particular that her taxonomy only partially uncovers the full causal diversity among female delinquents and that several latent classes are unidentified.

Purpose: This paper has four linked purposes: 1. To examine the strength of typological structure among female delinquents, 2. To estimate the number of distinct causal types that can be reliably identified, 3. To examine reliability and validity of these types and, 4. To relate these findings to the prior studies of female delinquent explanatory typologies.

Method: A sample of 1175 female delinquents from three State Juvenile Justice systems was assessed on a social-psychological battery of family, peer, school, community, lifestyle and personality factors. Official criminal histories were obtained. Pattern-seeking taxonomic methods were used to identify types (standard k-means, bootstrapped aggregation k-means and a semi-supervised network method). Cross-verification and external validation tests were applied.

Results: Seven reliable clusters emerged across replications - with kappa coefficients from 0.70 to 0.80 between pairs of replicated solutions. Analogues of Moffitt's AL and LCP types were found. However, several additional sub-types representing socially marginalized, neurotic internalizing (sexually abused) and "normal" low risk patterns were found. While reliable taxonomic structure was established, the boundaries between types were unstable and many hybrid cases exist.

Conclusions: While Moffitt's taxonomic categories are partially confirmed, several additional replicable patterns of female delinquents missing from her system are identified and described. These patterns appear to recur in several prior taxonomic studies of female delinquents. These findings may challenge the dominant theoretical paradigm that a single "general" causal process underlies all delinquency. The presence of "hybrid" cases suggests that blurred or fuzzy boundaries exist between clearly reliable and stable prototypes.

Key words:Female delinquents, taxonomy, patterns, cluster analysis, bootstrap aggregation