The present study examined the interrelationship of risk for recidivism, treatment readiness and responsivity, treatment-related change, treatment attrition, and recidivism among sexual offenders. It provided a cross validation of selected risk assessment measures, including the Static-99R, STABLE 2007, and Violence Risk Scale – Sexual Offender version (VRS-SO), as well as a clinical rating scale, the Treatment Readiness, Responsivity, and Gain Scale: Short Version (TRRG:SV). The primary focus of the study was to evaluate the relationship between treatment-related change and recidivism. The study was archival and participants included 185 federally incarcerated adult male sex offenders who participated in the Clearwater Sex Offender Treatment Program at the Regional Psychiatric Centre (Saskatoon, SK) between 1997 and 2001 and were followed up for an average of 9.3 years (SD = 3.0) post-release. Twenty percent of the sample was convicted for a new sexual offense, 45% for any new violent (including sexual) conviction, and 61% for any new conviction. The Static-99R, STABLE 2007, and VRS-SO all predicted violent recidivism (AUC = .62 to .72), as did the TRRG:SV (AUC = .32 to .37). Moreover, the dynamic risk measures demonstrated significant incremental validity, controlling for the Static-99R, in the prediction of sexual and violent recidivism. Treated participants made significant pre- to post-treatment changes on the VRS-SO, STABLE 2007, and TRRG:SV. Changes on the VRS-SO were significantly associated with lower rates of violent recidivism, with and without controlling for pre-treatment risk. Changes on the TRRG:SV were significantly associated with lower rates of sexual and violent recidivism. Significant predictors of treatment attrition were identified in the domains of criminal history, pre-treatment risk, treatment readiness and responsivity issues, and institutional adjustment. Implications for offender assessment, management, and rehabilitation are discussed.

Assessments designed to estimate the probability of sexual offense recidivism have gained popularity due to research that suggests these instruments are psychometrically sound for this purpose. The Static-2002 (Hanson & Thornton, 2003), an actuarial instrument of sexual recidivism, has been validated in specific populations, but in the absence of multicultural populations. It is necessary to establish the validity of the Static-2002 for a multicultural population, especially considering the increasing impetus to mandate measures of sexual recidivism. This study was designed to assess the predictive validity of the Static-2002 as an actuarial instrument that estimates the probability of sexual, violent, and general recidivism within a multicultural population. Recent criminal histories and de-identified archival records of 103 adult sex offenders were used to complete demographic questionnaires and to calculate retroactive predictive accuracy scores for the Static-2002. Records were obtained from a private practice office in South Florida, which provides outpatient treatment to conditionally released offenders. Hypothesis testing utilized a receiver operating characteristic area under the curve (ROC AUC) analysis to examine the discriminating power of the Static-2002 and to identify optimal cutoff values for each risk category. These cutoff values were then compared to the established Static-2002 risk categories to evaluate the cutoff scores' congruency. For this sample population, resulting ROC AUC values indicated the predictive accuracy of the Static-2002 to be moderate for any (including general; .67) offense recidivism, and high for violent recidivism (.74). However, the Static-2002 was ineffective for predicting sexual recidivism among this sample population since the predictive accuracy of the ROC area for sexual reoffenses was not statistically significant above chance (p = .07). The four Static-2002 risk categories derived for this study are: low, moderate-low, moderate-high, and high. When compared with the established Static-2002 risk categories, this study's risk categories were found to be mostly concurrent with the established categories. Similar to previous research, this study found that when the sample population differs racially/ethnically from the normed population (60% of this study's participants were Hispanic), the Static-2002 was ineffective for accurately predicting sexual reoffending. However, the Static-2002 effectively predicted violent and general recidivism with this sample population.

Over the last decade there has been a substantial research effort directed toward understanding recidivism and risk prediction among adult sex offenders. In contrast, the juvenile sex offender field has received much less empirical attention. Consequently, there are very few risk assessment instruments designed for use with young persons who have committed sexual offences. Available measures do not appear to take into account the higher rates of non-sexual recidivism typical of this population. This thesis aimed to identify risk factors for sexual and non-sexual recidivism among juvenile sex offenders (JSOs) and explore their utility in the context of a risk assessment instrument. In Study One, a meta-analysis was conducted on JSO research to examine risk factors for sexual recidivism. The meta-analysis involved 17 studies and 22 predictor constructs. Seven variables emerged as reliable predictors of sexual recidivism: an index sexual offence involving a stranger victim, a history of physical abuse in the offender's background, the presence of sexual deviance, a non-contact index sexual offence, sexually offending against a greater number of victims, a history of sexual offending, and a history of non-sexual offending. In Study Two, a meta-analysis was performed on JSO research to identify risk factors for non-sexual recidivism. The meta-analysis involved 14 studies and 19 predictor constructs. Six variables emerged as reliable predictors of non-sexual recidivism: a history of non-sexual offending, a history of sexual abuse (negative relationship), the presence of an antisocial orientation, an index sexual offence involving a child victim (negative relationship), a history of physical abuse in the offender's background, and the use of threats/force in the index sexual offence. Taken together, the two meta-analyses were able to identify risk factors specific to sexual recidivism (e.g., sexual deviance) and non-sexual recidivism (e.g., antisocial orientation) as well as risk factors common to both outcomes (e.g., a history of non-sexual offending). These studies therefore made an original and significant contribution to the understanding of recidivism among JSOs. Based upon the meta-analyses, Study Three involved the construction of a risk assessment checklist that could be easily scored from file information. The Juvenile Risk Assessment Checklist (J-RAC) contained 11 items based upon the most consistent operationalisations of the predictors that emerged from the meta-analyses. The J-RAC is unique in the JSO risk assessment field as it contains two scales designed to provide separate estimates of the risk of youth engaging in further sexual and non-sexual recidivism, respectively. The J-RAC was scored based on file information from 82 youth who had been found guilty of sexual offences and referred to a treatment service in Queensland, Australia. The majority of the items possessed good inter-rater reliabilities and the two scales both possessed high inter-rater reliabilities. An existing JSO risk assessment instrument, the Juvenile Sex Offender Assessment Protocol Version 2 (J-SOAP-II; Prentky & Righthand, 2003), was also administered and the J-RAC possessed significant correlations with the majority of the J-SOAP-II scales. The J-RAC was also found to significantly discriminate between juveniles in detention and juveniles in the community, with those in detention found to be at higher risk on both scales. Thus, the reliability, convergent validity, and discriminant validity of the J-RAC was established. Consistent with current adult sex offender recidivism theory, the findings demonstrated the distinct roles of sexual deviance and antisocial orientation in the prediction of sexual and non-sexual recidivism of JSOs, respectively. The current thesis goes some way towards advancing a theory of juvenile sex offender recidivism by drawing upon the variables that emerged as reliable risk factors in the two meta-analyses. While the J-RAC could not yet be considered an actuarial instrument capable of assigning low, medium and high risk levels, the results of the research conducted in this thesis are encouraging. The utility of the J-RAC may be further enhanced through the aid of prospective research that incorporates substantial numbers of juvenile recidivists.

Le présent projet doctoral vise à considérer les lacunes dans la documentation scientifique sur le Trouble Paraphilique Coercitif (TPC) en mettant l’accent sur la validité des critères diagnostiques proposés pour inclusion dans le DSM-5 et les marqueurs comportementaux. À ce fait, les données archivées d’individus ayant sexuellement agressé des femmes adultes ont été étudiées. La thèse est constituée de trois articles empiriques. Le premier article présente des résultats clés découlant des analyses, élaborés dans les articles subséquents. Le second (N = 47) évalue les fréquences observées du TPC, la validité et l’impact du recours au nombre minimal de victimes comme critère diagnostique, ainsi que les indices prédisant la récidive sexuelle. Le troisième article (N = 52) compare les groupes diagnostiques sur une série de comportements délictuels, tels que les gestes sexuels et les comportements violents, dans le but d’identifier les marqueurs comportementaux associés avec la propension au viol qui pourraient assister dans le processus diagnostique. Dans le même ordre d’idées, nous avons créé des typologies de violeurs à partir des gestes sexuels commis, d’un côté, et des comportements violents, de l’autre côté. Conséquemment, les caractéristiques des typologies ainsi obtenues et leur association avec le TPC furent examinées. Dans l’ensemble, nos résultats ne soutiennent pas le recours au nombre de victimes. Nos données suggèrent que, globalement, les violeurs avec le TPC utilisent un niveau de gestes sexuels plus envahissant et un niveau de violence moindre que les violeurs n’ayant pas ce diagnostic, et que l’exhibitionnisme et l’attouchement pourraient servir de marqueurs comportementaux pour le TPC. En outre, les violeurs avec le TPC sont caractérisés davantage par demande indécente, exhibitionnisme, attouchement, masturbation, tentative de pénétration et pénétration digitale que par pénétration vaginale et sodomie. De plus, ces derniers font moins recours à l’utilisation d’armes, semblent ne pas frapper/donner des coups à la victime et sont caractérisés par la manipulation plutôt que par le recours aux menaces de mort, force excessive et utilisation d’armes. En somme, nos données soulignent la nécessité de s’appuyer sur une combinaison de méthodes d’évaluation afin d’améliorer la validité diagnostique et discriminante du TPC. / The present dissertation aims to address the shortcomings in the current literature on Paraphilic Coercive Disorder (PCD) by focusing on two main objectives: assessing the validity of the diagnostic criteria proposed for inclusion in the DSM-5 and investigating behavioral markers. To this end, archival files of rapists who offended against adult women were studied. The thesis consists of three empirical articles. The first article presents a succinct account of some of the key results emanating from the analyses. The second article (N = 47) examines the observed frequencies of PCD and assesses the validity and impact of relying on minimum number of victims as a diagnostic criterion. Furthermore, a number of variables of interest are examined to determine predictors of sexual recidivism. The third article (N = 52) compares diagnostic groups on a number of offense conduct characteristics – specifically sexual acts and violent behaviors – in an attempt to identify behavioral markers associated with rape-proneness that could aid with the diagnosis of PCD. Similarly, rapist typologies were created by classifying the sample into groups of sex offenders based on their sexual acts, on one hand, and violent behaviors, on the other hand. Consequently, their characteristics and association with PCD were examined. Our results do not support the reliance on number of victims. Our findings suggest that rapists with PCD are more sexually intrusive and resort to less violence overall than sex offenders without such a diagnosis and that exhibitionism and fondling could serve as behavioral markers for PCD. Moreover, rapists with PCD are characterised more by indecent request, exhibitionism, fondling, masturbation, attempted intercourse and digital penetration rather than by intercourse and sodomy. In terms of violent behaviors, rapists with PCD resort less to the use of weapons, seem not to hit their victims, and are likely characterised more by manipulation rather than by the use of death threats, excessive force and weapons. In sum, the present study highlights the necessity of relying on a combination of assessment methods in order to improve diagnostic and discriminant validity of PCD.