A cluster analytic examination and validation of adult victim sexual offending subtypes in two Canadian samples

• Published in: Sexual offending Vol. 16
• Main Authors: Myburgh, John-Etienne, Olver, Mark E.
• Format: Journal Article
• Language: English
• Published: PsychOpen GOLD/ Leibniz Institute for Psychology 14.12.2021
• Subjects: model-based cluster analysis; risk assessment; sexual offense typology; static 99r; vrs-so
• ISSN: 2699-8440; 2699-8440
• DOI: 10.5964/sotrap.3741

The development and validation of sexual offense perpetrator typologies remains a useful endeavor with implications for theory and correctional/clinical practice. Most such typologies—which rely on factors such as the individual’s motivation for offending—have not been validated empirically. The current study utilized a validated sexual violence risk-needs instrument, the Violence Risk Scale—Sexual Offense version (VRS-SO; Wong, Olver, Nicholaichuk, & Gordon [2003, 2017], Regional Psychiatric Centre and University of Saskatchewan, Saskatoon, Canada), to develop and validate an empirically-derived adult victim sexual offense (AVSO) typology through model-based cluster analysis of dynamic risk-need domains. The study featured two treated samples of men (n = 283 and 169) convicted for contact sexual offenses against adult victims. A three-cluster solution was identified and replicated across the two samples: high antisociality high deviance (HA-HD), high antisociality low deviance (HA-LD), and low antisociality low deviance (LA-LD). External validation analyses demonstrated that HA-HD men had more dense sexual offense histories, were more likely to be diagnosed with a paraphilia, and had the highest rates of sexual recidivism (Sample 2 only). By contrast, the HA-LD men had greater concerns on indexes of nonsexual criminality, particularly high base rates of antisocial personality and substance use disorders, and high rates of general violent recidivism (particularly Sample 1). The findings suggest that the VRS-SO factors may have utility in discriminating between AVSO types to inform sexual offending theory, case formulation, and risk management. Background Theory and research have well documented that men who commit acts of sexual aggression toward women vary on important psychological, historical, and offence-related dimensions, and that these men can be grouped into different categories or subtypes of perpetrator. The present study examined the possible presence of subtypes of men who target women for sex crimes who may be grouped on the basis of psychological risk factors. Why was the study done? The existence of a reliable typology can be useful to inform rehabilitation and risk management methods to prevent further victimization and to allow the safe reintegration of these men to society. What did the researchers do and find? Two Canadian samples of men who were incarcerated for sexual assaults against adult females were examined on the basis of the psychological risk factors measured by a test called the Violence Risk Scale-Sexual Offense version (VRS-SO). These men had also attended treatment for sexual offending. The same three subtypes emerged in each sample. One group was high on problematic sexual interests and behaviors as well as being rule violating and criminalized; they also held problematic attitudes toward sex crimes. A second group was highly criminalized, had a significant history of nonsex crimes, and held negative attitudes, but they were not sexually deviant. A third group was low on all three dimensions (sexual functioning, criminality, attitudes and responsivity). What do these findings mean? The results are consistent with other findings of subtypes of men who sexually assault women and suggest that some sex crimes may be driven by poor management of sexual interests and behaviors, while others may reflect broader patterns of lifestyle criminality and problems in emotional functioning. These findings have implications about where to invest treatment resources and which individuals may be most risky to reoffend violently or sexually. Highlights Model based clustering of VRS-SO dynamic factor scores generated three subtypes that were replicated across treated samples of men with adult victims of sexual offenses (AVSO): high antisociality-high deviance (HA-HD), high antisociality-low deviance (HA-LD), and low antisociality-low deviance (LA-LD). The findings are consistent with empirical research examining AVSO subtypes from other structured classification systems (e.g., Massachusetts Treatment Center typologies). HA-HD men tended to be older, have more extensive histories of sexual offending, higher rates of paraphilia, and (Sample 2) higher rates of sexual recidivism. HA-LD men tended to be younger, with extensive histories of nonsexual criminality, high base rates of substance use disorder, and (Sample 1) higher rates of violent recidivism. LA-LD men were lowest risk across the VRS-SO domains examined and had the lowest rates of recidivism, despite high base rates of antisocial personality and substance use disorders.