A comparison of modified versions of the Static-99 and the Sex Offender Risk Appraisal Guide

The Effect of Missing Item Data on the Relative Predictive Accuracy of Correctional Risk Assessment Tools

Missing data are pervasive in risk assessment but their impact on predictive accuracy has largely been unexplored. Common techniques for handling missing risk data include summing available items or proration; however, multiple imputation is a more defensible approach that has not been methodically tested against these simpler techniques. We compared the validity of these three missing data techniques across six conditions using STABLE-2007 (N = 4,286) and SARA-V2 (N = 455) assessments from men on community supervision in Canada. Condition 1 was the observed data (low missingness), and Conditions 2 to 6 were generated missing data conditions, whereby 1% to 50% of items per case were randomly deleted in 10% increments. Relative predictive accuracy was unaffected by missing data, and simpler techniques performed just as well as multiple imputation, but summed totals underestimated absolute risk. The current study therefore provides empirical justification for using proration when data are missing within a sample.

The Council of State Governments Justice Center Approach to Increasing Risk-Level Consistency in the Application of Risk Assessment Instruments

Correctional and forensic mental health settings potentially have multiple risk assessment instruments administered on a single client. Because of the various methods of determining risk categories, risk-level consistency can become an issue. The Council of State Governments Justice Center developed a Five-Level System that can be applied to most risk assessment instruments. Using the Level of Service Inventory-Revised and two created risk assessment instruments, the present study assessed if the Five-Level System (vs. normative percentile categories) demonstrated greater agreement between the two instruments, and, if so, the percentage of greater agreement. The Five-Level System demonstrated 4% to 5% greater agreement for both risk-level placement and recidivism rates. The implications of this greater consistency among risk assessment instruments is an increased fairness in making risk-level assignments.

The Role of Psychopathic Traits in the Development of the Therapeutic Alliance Among Sexual Offenders

There is a growing body of research demonstrating that the therapeutic alliance (TA) affects outcomes among specialized forensic populations, including sexual offenders. Despite this consensus, researchers continue to question whether higher levels of psychopathic traits are conducive to the formation of a therapeutic relationship for high-risk sexual offenders. Thus, the current study adds to the literature by examining the relationship between the TA and levels of psychopathy among a sample of incarcerated sexual offenders participating in sexual offender treatment. Overall, we found no significant relationships between Psychopathy Checklist-Revised (PCL-R) scores and the Working Alliance Inventory (WAI) for either client or therapist ratings. However, when we excluded those offenders who were participating in aftercare, a significant negative relationship was found between client ratings of the Bonds subscale and PCL-R total scores. Next, after controlling for risk and group status (aftercare/non-aftercare), we found no significant differences between either client or therapist total WAI scores when compared by level of psychopathy as measured by the PCL-R (low, >20; moderate, 20-30; and high, >30). Furthermore, when Factor 1 and Factor 2 scores of the PCL-R were examined individually, neither factor significantly predicted either client or therapist total WAI score after controlling for risk and group status. Findings are discussed as they pertain to the treatment of sexual offenders with elevated levels of psychopathic traits.

Challenges in the treatment of dangerous and severe personality disorder

Dangerous and severe personality disorder (DSPD) services have a relatively short history but are increasingly involved in the implementation of therapeutic programmes. We describe the background to the DSPD initiative in England and consider issues that arise in planning and delivering treatment services. Two bodies of evidence are particularly relevant: previous research into personality disorder and its treatment, which we suggest is as yet of limited value, and research into the outcomes of offender treatment programmes. The latter is clearly relevant but greater consideration of adapting programmes for the patient population and of breadth of treatment is required in the DSPD setting. The important task is to integrate components for the treatment of personality disorder and offending behaviour in a holistic manner. We describe three further challenges in delivering treatment and suggest that ongoing evaluation of treatments is critical in this area of practice, given the impoverished knowledge base.

Sexually violent predator risk assessments with the violence risk appraisal guide-revised: A shaky practice

Twenty-two jurisdictions in the United States permit the involuntary civil confinement of sexual offenders upon expiration of their criminal sentence and, if committed, these individuals face possible lifetime commitment. One of the legal requirements that psychologists must address in sexually violent predator evaluations is the likelihood that an individual will engage in dangerous sexual behavior and consideration of the probabilities for sexual recidivism contained in actuarial experience tables best address this inquiry. Clinicians find it increasingly difficult to affirm the likelihood threshold in the face of decreasing base rates and score-wise probability estimates for sexual recidivism reported in contemporary actuarial experience tables. The Violence Risk Appraisal Guide-Revised (VRAG-R) has been promoted to assess sexually violent predators because it has been presented as a more accurate predictor of sexual recidivism and the results more likely satisfy the legal standard of sexual dangerousness. This article conducts an in-depth analysis of the predictive and psychometric properties of the VRAG-R that are most relevant to the fit of the VRAG-R when addressing the sexual dangerousness standard proscribed by SVP laws. Recommendations for future research are offered to improve the fit of the VRAG-R to the legal inquiry of sexual dangerousness and implications for using the current iteration of the VRAG-R in forensic practice are discussed.

A Literature Review of the Utility of Selected Violence and Sexual Violence Risk Assessment Instruments

The rationale for this article is to comprehensively outline and describe the strengths and weaknesses of various risk assessment instruments/tools relevant to the evaluation of sexually violent and violent offenders. The author will briefly discuss ethical obligations the forensic mental health professional (FMHP) must consider when conducting risk assessments.

Sexual Offenders' Perceptions of the Client-Therapist Relationship: The Role of Risk

The therapeutic alliance has been shown to be integral to treatment outcomes even in forensic settings. There is still a relative dearth of research examining factors related to the formation of the therapeutic alliance in sex offender treatment specifically. Using a sample of 202 incarcerated male sexual offenders participating in sex offender treatment, this study examined whether perceptions of the client-therapist relationship from the perspective of male sexual offenders varied by risk for sexual and general recidivism. Overall, we found a significant negative relationship between risk for sexual recidivism and bond formation. However, when therapist subscale scores on the Working Alliance Inventory were considered by therapist gender, higher risk sexual offenders perceived poorer bonds with their female therapists, relative to their male therapists. Findings are discussed as they pertain to therapeutic relationships and responsivity issues in sex offender treatment.

The Westgate Service and Related Referral, Assessment, and Treatment Processes

The formerly named "Dangerous and Severe Personality Disorder" (DSPD) units are no longer standalone services within the criminal justice system in England and Wales. These sites now provide personality disorder treatment services in the high-security prison estate as part of the new national Offender Personality Disorder (OPD) Pathway Strategy. The OPD Pathway intends to take responsibility for the assessment, treatment, and management of offenders who are likely to have a personality disorder and who present a high risk of re-offending (men and women) and serious harm to others (men). Further PD treatment and progression services are being commissioned in lower security prisons and in the community as part of the new PD Strategy. While the suitability criteria for the two male high-security PD treatment sites are the same, the individual units have their own assessment and treatment methods. This article aims to communicate the referral, assessment, and treatment methods employed within the prison-based Westgate Personality Disorder Treatment Service, HMP Frankland.

A multimodal examination of sexual interest in children: a comparison of sex offenders and nonsex offenders

Research and theoretical models have consistently identified sexual interest in children as a key factor involved in child sexual offending. However, there is only moderate agreement in the diagnosis of pedophilia and different assessment methods identify different offenders as pedophiles. The current study examined the discriminative and convergent validity of three different measures of sexual interest in children. Participants included sex offenders and nonsex offenders recruited from federal prisons (i.e., offenders serving sentences of more than 2 years) in Ontario, Canada. Child molesters' responses (n = 35) were not significantly different from nonsex offenders (n = 21) on an implicit measure of sexual interest in children (Sexual Attraction to Children Implicit Association Test [SAC-IAT] d = 0.44, 95% CI [-0.11, 0.99]), but differed on the self-report (Sexual Interest Profiling System; d = 0.83, 95% CI [0.27, 1.39]) and viewing time (d = 1.15, 95% CI [0.54, 1.75]) measures. Findings did not provide clear support for the superiority of a multimodal approach, possibly due to the relatively small sample. More often than not, convergence between the three measures was observed (n = 74). Findings from the present study are an important step toward understanding the relationship between different measures of sexual interest in children and establishing their validity.

An exploration of crossover sexual offending

Studies have produced equivocal findings regarding whether sex offenders are stable in their choice of victims. Indeed, it remains unclear whether a sex offender's subsequent victims are typically of the same gender, age range, and victim-perpetrator relationship as that of the initial victim. Although some differences may be attributed to methodological disparities, others are not. This study sought to clarify this question by examining the tendency of sex offenders to switch from one type of victim to another, both within an index offense and across offenses and all victims. Archival records of 789 incarcerated sex offenders were examined. Of those offenders who had multiple victims at the index offense (n = 279), 13% had victims of both genders, 14% had victims in different age categories (child, adolescent, and adult), and 13% had varying relationships with the victims (i.e., family member, acquaintance, or stranger). When the records of those with past sexual convictions were examined (n = 208), 20% of offenders had a prior victim of a different gender; 40% crossed over across age categories, and 48% of the repeat offenders had varying relationships with the victim across convictions. Offenders who had both male and female victims and offenders who had victims of varied relationship status across crimes had higher Static-99 risk scores than offenders who were more stable with regard to victim selection. These findings are compared to those of previous studies, focusing on how these results add clarity to a previous literature whose conclusions were challenged by the use of disparate sampling and research methodologies.

Does volunteering for sex offender treatment matter? Using propensity score analysis to understand the effects of volunteerism and treatment on recidivism

A common critique of program evaluations of prison-based sex offender treatment holds that the samples inherently show selection bias because the participants typically volunteer for treatment. To address this critique, we used propensity score analysis to assess the influence of volunteerism on treatment effects. We examined recidivism outcomes for a sample of participants who volunteered for treatment, of whom some participated in treatment (n = 161) and some did not (n = 282) and compared these outcomes to the recidivism rate of a matched sample of nonvolunteers for treatment (n = 443). The primary finding is that offenders who volunteered for treatment did not demonstrate any differences in recidivism rates when matched with and compared to inmates who did not volunteer to participate in treatment. Furthermore, our results revealed that there were a number of significant differences between unmatched volunteers and unmatched nonvolunteers, perhaps most importantly in their risk for future recidivism as measured by the STATIC-99 risk assessment. We discuss study strengths and limitations and present the implications of the findings for policy, practice, and research.

An examination of suicide attempts among incarcerated sex offenders

Little is known about suicide attempts among sex offenders. This study examines the rates of nonfatal suicide attempts among a sample (N = 3,030) of incarcerated male sex offenders. Overall, the authors found that 14% of sex offenders in the study sample had made a suicide attempt at some point in their lives. Of those, 11% had reported a suicide attempt prior to incarceration, 0.5% had made a suicide attempt while incarcerated, and 2.5% made suicide attempts both prior to and during incarceration. Sex offenders who made suicide attempts were significantly more likely than those who did not make suicide attempts to have had an abusive childhood, a history of psychiatric problems, intellectual impairment, male victims, and related victims. Suicide attempters also scored higher on actuarial risk measures than nonattempters. No differences were found in attempter status between sex offenders who committed sex offenses against children and those who committed sex offenses against adults. A history of psychiatric problems and treatment as well as childhood abuse/neglect and perpetration against male victims predicted suicide attempter status. These findings are discussed as they pertain to suicide prevention, risk assessment, and the collateral consequences of sex offender legislation.

Are actuarial risk data used to make determinations of sex offender risk classification? An examination of sex offenders selected for enhanced registration and notification

This study examined whether evaluators use actuarial risk scores and risk information to make determinations about sex offender risk status for the purpose of enhanced registration and notification. Although it was expected that sexual offenders selected for enhanced registration and notification would have higher scores on actuarial risk assessment tools than those who were not selected, few differences were found between groups with regard to risk factors associated with sexual offense recidivism. Given that actuarial tools enhance the prediction of sexual recidivism, this study may shed light on problems in the implementation of sex offender policy measures. Results are discussed as they pertain to the assessment and application of registration and community notification statutes for sexual offenders.

Multidimensional assessment of criminal recidivism: problems, pitfalls, and proposed solutions

All states have statutes in place to civilly commit individuals at high risk for violence. The authors address difficulties in assessing such risk but use as an example the task of predicting sexual violence recidivism; the principles espoused here generalize to predicting all violence. As part of the commitment process, mental health professionals, who are often psychologists, evaluate an individual's risk of sexual recidivism. It is common for professionals conducting these risk assessments to use several actuarial risk prediction instruments (i.e., psychological tests). These tests rarely demonstrate close agreement in the risk figures they provide. Serious epistemological and psychometric problems in the multivariate assessment of recidivism risk are pointed out. Sound psychometric, or in some cases heuristic, solutions to these problems are proffered, in the hope of improving clinical practice. The authors focus on how to make these tests' outputs commensurable and discuss various ways to combine them in coherent, justifiable fashions.

A time series modeling approach in risk appraisal of violent and sexual recidivism

For over half a century, various clinical and actuarial methods have been employed to assess the likelihood of violent recidivism. Yet there is a need for new methods that can improve the accuracy of recidivism predictions. This study proposes a new time series modeling approach that generates high levels of predictive accuracy over short and long periods of time. The proposed approach outperformed two widely used actuarial instruments (i.e., the Violence Risk Appraisal Guide and the Sex Offender Risk Appraisal Guide). Furthermore, analysis of temporal risk variations based on specific time series models can add valuable information into risk assessment and management of violent offenders.

Prospective actuarial risk assessment: a comparison of five risk assessment instruments in different sexual offender subtypes

This study examines the predictive validity of the most commonly used risk assessment instruments for sexual offenders: Static-99, Rapid Risk Assessment for Sexual Offense Recidivism, Sex Offender Risk Appraisal Guide, Sexual Violence Risk-20, and Psychopathy Checklist-Revised in a prospective research design. Although risk assessment is part of a regime leading to various efforts to reduce risk by treatment and aftercare, all instruments show good predictive validity. However, depending on the instrument, recidivism category, and subgroup, the predictive accuracy varies markedly. Furthermore, the authors fail to demonstrate predictive validity for sexual violent reoffences-for the whole sample and for all subgroups. The results, nevertheless, support the utility and predictive validity of actuarial risk assessment complementary to treatment efforts to reduce risk. On the other hand, forensic practitioners have to be aware of the limitations of actuarial risk assessment methods, in particular as regards to variable predictive accuracy for different sexual offender subgroups and reoffence categories.

Family experiences of young adult sex offender registration

Since 1994, policies have been instituted throughout the United States that require sex offenders to register their personal information with law enforcement officials (Jacob Wetterling Crimes against Children and Sexually Violent Offender Registration Program, 1994). With the passage of additional laws, this information is now available to the public via the Internet or a request to a police department. These laws have brought about consequences for both the registrants and for members of their families. A focus group was held with four mothers who have sons listed on the Michigan Sex Offender Registry. Psychological and social consequences of registration were found and policy implications are discussed.

A descriptive profile of high-risk sex offenders under intensive supervision in the province of British Columbia, Canada

This study provides a preliminary descriptive profile of individuals having been issued an 810 recognizance order (i.e., peace bond). This preventive order is issued to individuals in the community considered by the court to be high-risk sex offenders. In total, 88 offenders were issued an 810.1 (at risk of sexual offence against a child) or an 810.2 (at risk of a personal injury offence) recognizance order between April 2006 and February 2008, among whom 59 sex offenders were included. The sociodemographics, risk profiles, and recidivism of offenders under the orders were analyzed. Two risk assessment tools, the Static-99 and the Stable, were used to describe this population. Comparisons were made between sex offenders under 810.1 and 810.2 orders. Preliminary analyses suggest that these men are heterogeneous in terms of their level of risk of reoffending. Only one offender sexually reoffended while under the order. On the other hand, the general recidivism rate was about 30%, with many breaching the conditions of their order. The findings are discussed in light of the community risk management of high-risk sex offenders.

Accuracy of the static-99 in predicting recidivism in Switzerland

The Static-99 is a widely used actuarial risk assessment instrument. Various international validation studies have found satisfactory to good predictive validity for the Static-99, with the area under the curve (AUC) between 59% and 95%. This study is the first evaluation of the predictive accuracy of the Static-99 among sex offenders in Switzerland. The Static-99 scores of 69 violent/sex offenders in Switzerland were assessed using data from their psychiatric assessments. Recidivism was operationalized as reconviction assessed from penal records. The Static-99 risk levels were predictive for recidivism (AUC = .758) among our population. The results are discussed on the basis of the literature.

Sexual offenders in preventive detention: data concerning the inmates and expert witness practice

In the last decade, preventive detention-especially that relating to sexual offenders- has gained relevance for the German legal system. However, data are lacking concerning the inmates and the modus operandi of the psychiatric experts. Court orders and psychiatric statements of 114 offenders with orders of preventive detention were examined, 57 of whom were incarcerated because of sexual offences. Sexual offenders mostly show antisocial personality traits or even disorders, with a history of poly-trophic delinquency. In only four cases, the diagnosis of a sexual disorder was given. The analysis showed a high rate of psychiatric expert testimonies' lacking basic information, with incomplete assessment of sexual history. None of the experts used standardized prognostic instruments, meaning that most of the risk factors included in the Sexual Violence Risk-20 and the Static-99 were not considered. Further work needs to be done to improve the quality of psychiatric statements concerning sexual offenders.

Long-term predictive validity of the risk matrix 2000: a comparison with the static-99 and the sex offender risk appraisal guide

The purpose of this study is to examine the predictive accuracy of the Risk Matrix 2000 on an independent sample of 351 sexual offenders, followed in the community for an average duration of 11.4 years (range 0-20 years, SD = 4.4 years). For comparison purposes, this study also examines the predictive accuracy of two other risk assessment instruments, specifically modified versions of the Static-99 and the Sex Offender Risk Appraisal Guide (SORAG). Results indicate that the Risk Matrix 2000 demonstrates convergent validity by correlating with the other risk assessment instruments. Moreover, the Risk Matrix 2000 is predictive of recidivism above chance levels, exhibiting medium to large effect sizes, although in general, the other two instruments, particularly the SORAG, are superior. Results also indicate differences in predictive validity when comparing 2-year, 5-year, and variable follow-up periods. Finally, a cumulative meta-analysis compares and integrates current findings with those obtained from the accumulation of previous studies, and new cumulative estimates are provided.

Pornography use and sexual aggression: the impact of frequency and type of pornography use on recidivism among sexual offenders

In this study, we examined the unique contribution of pornography consumption to the longitudinal prediction of criminal recidivism in a sample of 341 child molesters. We specifically tested the hypothesis, based on predictions informed by the confluence model of sexual aggression that pornography will be a risk factor for recidivism only for those individuals classified as relatively high risk for re-offending. Pornography use (frequency and type) was assessed through self-report and recidivism was measured using data from a national database from the Royal Canadian Mounted Police. Indices of recidivism, which were assessed up to 15 years after release, included an overall criminal recidivism index, as well as subcategories focusing on violent (including sexual) recidivism and sexual recidivism alone. Results for both frequency and type of pornography use were generally consistent with our predictions. Most importantly, after controlling for general and specific risk factors for sexual aggression, pornography added significantly to the prediction of recidivism. Statistical interactions indicated that frequency of pornography use was primarily a risk factor for higher-risk offenders, when compared with lower-risk offenders, and that content of pornography (i.e., pornography containing deviant content) was a risk factor for all groups. The importance of conceptualizing particular risk factors (e.g., pornography), within the context of other individual characteristics is discussed.

Sexual reconviction rates in the United Kingdom and actuarial risk estimates

OBJECTIVE

Assessing the risk of further offending behavior by adult sexual perpetrators of children is highly relevant and important to professionals involved in child protection. Recent progress in assessing risk in sexual offenders has established the validity of actuarial measures, although there continues to be some debate about the application of these instruments. This paper summarizes the debate between clinical and actuarial approaches and reviews the "base rate" for United Kingdom sexual offense reconviction.

METHOD

A review of the literature revealed 16 UK sexual reconviction studies, 8 using incarcerated samples (N=5,915) and 8 using non-incarcerated samples (N=1,274). UK estimates of sexual reconviction rates are compared with European and North American studies.

RESULTS

The mean sexual reconviction rates for the incarcerated sample at 2 years (6.0%), 4 years (7.8%) and 6 years or more (19.5%) were higher than that of the comparative non-incarcerated sample at 2 years (5.7%), up to 4 years (5.9%), and 6 years or more (15.5%). The overall sexual reconviction rate for both samples combined was 5.8% at 2 years, and 17.5% at 6 years or more.

CONCLUSIONS

The sexual reconviction rate for incarcerated sexual offenders is higher than that of non-incarcerated sexual offenders. The UK sexual reconviction rates were comparable with European and North American studies.

Incarceration and recidivism among sexual offenders

The relationship between incarceration and recidivism was investigated in a sample of 627 adult male sexual offenders. Incarceration for the index offense was unrelated to sexual or violent recidivism. This was the case whether incarceration was examined as a dichotomous variable (incarceration vs. community sentence) or as a continuous variable (length of incarceration). Risk for sexual recidivism was assessed with a modified version of the Rapid Risk Assessment for Sexual Offense Recidivism. There was no evidence that the relationship between incarceration and recidivism was confounded or moderated by risk or that length of incarceration and recidivism were non-linearly associated. Sentencing sexual offenders to terms of incarceration appears to have little, if any, impact on sexual and violent recidivism following release.

Denial predicts recidivism for some sexual offenders

This study examined whether there were variables that moderated the relationship between denial and recidivism among adult male sexual offenders. The first study (N = 489) found that the relationship with sexual recidivism was moderated by risk (as measured by the Rapid Risk Assessment for Sexual Offense Recidivism) but not by psychopathy (as measured by the Psychopathy Checklist-Revised). Contrary to expectations, denial was associated with increased sexual recidivism among the low-risk offenders and with decreased recidivism among the high-risk offenders. Post hoc analyses suggested that the risk item most responsible for the interaction was "relationship to victims". For incest offenders, denial was associated with increased sexual recidivism, but denial was not associated with increased recidivism for offenders with unrelated victims. These interactions were substantially replicated in two independent samples (N = 490 and N = 73). The results suggest that denial merits further consideration for researchers as well as those involved in applied risk assessment of sexual offenders.

The validity of the Violence Risk Appraisal Guide (VRAG) in predicting criminal recidivism

INTRODUCTION

The VRAG is an actuarial risk assessment instrument, developed in Canada as an aid to estimating the probability of reoffending by mentally ill offenders.

AIM

To test the predictive validity of the VRAG with a German sample.

METHOD

The predictive validity of the VRAG was tested on a sample of 136 people charged with a criminal offence and under evaluation for criminal responsibility in the forensic psychiatry department at the University of Munich in 1994-95. The predicted outcome was tested by means of ROC analysis for correlation with the observed rate of recidivism between discharge after the 1994-95 assessment and the census date of 31 March 2003. Recidivism rate was calculated from the official records of the National Conviction Registry.

RESULTS

Just over 38% of the sample had reoffended by 2003. Their mean time-at-risk was 58 months (SD 3.391; range 0-115 months). The VRAG yielded a high predictive accuracy in the ROC analysis with an AUC of 0.703. For a constant time-at-risk < = 7 years, the predicted probability and observed rates of recidivism correlated significantly with Pearson's r = 0.941.

CONCLUSIONS

The validity of the VRAG was replicated with a German sample. The VRAG yielded good predictive accuracy, despite differences in sample and outcome variables compared with its original sample.

Cross-validation of the risk matrix 2000 sexual and violent scales

The predictive accuracy of the newly developed actuarial risk measures Risk Matrix 2000 Sexual/Violence (RMS, RMV) were cross validated and compared with two risk assessment measures (SVR-20 and Static-99) in a sample of sexual (n= 85) and nonsex violent (n= 46) offenders. The sexual offense reconviction rate for the sex offender group was 18% at 10 years follow-up, compared with 2% for the violent offenders. Survival analyses revealed the violent offenders were reconvicted at twice the rate compared to sexual offenders. The RMV significantly predicted violent recidivism in the sex and combined sex/violent offender groups. Although the RMS obtained marginal accuracy in predicting sexual reconviction in the sex offender group, none of the scales significantly predicted sexual reconviction. An item analysis revealed four factors not included in the risk scales that were significantly correlated with sexual and violent reconviction. Combining these factors with Static-99, RMV, and RMS increased the accuracy in predicting sexual reconviction.

Comparison of two risk assessment instruments for sexual offenders

The current research examines the predictive validity of the Static-99 and the SORAG in predicting sexual and violent recidivism among a sample of 258 treated high-risk sexual offenders. While the SORAG was found to have moderate predictive accuracy for both sexual and violent recidivism over a 5-year follow-up period, the Static-99 was found to only predict sexual recidivism. As well, the actual recidivism rates in the current sample were compared to the published risk percentages for each of the instruments. For both the Static-99 and the SORAG the current sample re-offended at a lower rate than expected. Possible explanations for this finding are discussed.

Evaluation of the SORAG and the Static-99 on Belgian sex offenders committed to a forensic facility

This study evaluated the convergent and predictive validity of the Sex Offender Risk Appraisal Guide (SORAG; V. L. Quinsey, G. T. Harris, M. E. Rice, & C. A. Cormier, 1998) and the Static-99 (R. K. Hanson & D. Thornton, 1999, 2000) among 147 male sex offenders committed to a high-security hospital in Belgium (Centre de Défense Sociale "Les Marronniers"). Of the sample, 63.8% were child abusers (victims under 14 years of age), 24.6% were rapists (victims aged 14 years or more), and 11.5% were mixed-victim offenders (victims less than 14 years of age and victims aged 14 years or more). After an average follow-up period of 4.2 years (SD = 3.4 years), the sexual recidivism rate was 25.2%, the violent recidivism rate was 17.1%, and the general (any) recidivism rate was 33.1%. The SORAG and the Static-99 were moderately correlated with each other (r = .55), and both showed strong predictive validity for general and violent recidivism (ROC AUC's ranging from .68 to .72 for the total sample). Both instruments showed moderate predictive validity for sexual recidivism (AUC of .64 for SORAG and .66 for Static-99).

Another look at interpreting risk categories

Several studies over the past decade have shown that simple rating scales can accurately rank sex offenders' long-term risk of recidivism. But when using these scales as prediction tools, evaluators often wish to translate categories of risk into probabilities of recidivism. D. M. Doren (2004) has recently suggested that evaluators may use the recidivism percentages published in original studies of the RRASOR and STATIC-99 without regard to differences in populations or base rates. This article explains why Doren's computations should lead to a different conclusion, and describes how simply comparing percentages across studies can mislead researchers and clinicians. Instead, investigators should isolate and examine the detection properties of risk assessment instruments alone, independent of the population- or setting-specific base rate. This article explains this process, using an imaginary study to illustrate how base rates and the properties of risk assessment instruments yield estimated probabilities of recidivism. The article also shows why Doren's results imply that the percentages of recidivism associated with scores on the RRASOR and STATIC-99 scores may vary across study populations. The article offers recommendations for researchers who design and evaluate actuarial methods of assessing risk and for clinicians who interpret results from risk assessment instruments.

Hostility and recidivism in sexual offenders

In this study, we examined the association of hostility, as measured by the Buss-Durkee Hostility Inventory (BDHI), with offence characteristics and recidivism in 656 adult male sexual offenders. Hostility was significantly associated with having prior violent charges, the use of violence in the index sexual offence, sexual recidivism, and violent recidivism. After controlling for risk level, as measured by a modified version of the Rapid Risk Assessment for Sexual Offence Recidivism (RRASOR-mod), the significant association between hostility and sexual and violent recidivism remained. When examined by type of offender, hostility was significantly associated with recidivism in intrafamilial and extrafamilial child molesters, but not in rapists or mixed offenders. Given the predictive value of hostility independent of the RRASOR-mod, the present findings confirm and encourage treatment efforts directed toward the management of hostility and anger in sexual offenders.

Toward a multidimensional model for sexual recidivism risk

The paradigm underlying current sexual offender recidivism risk assessment procedures conceptualizes such risk in a linear fashion, ranging on a single continuum from 0% to 100%. Each risk and protective characteristic thought of relevance in an evaluation is used as an indicator of increased or decreased risk, respectively, along that same continuum. This conceptualization of risk was useful as a starting place for the application of empirically supported risk and protective factors. On the other hand, this perspective is likely simplistic and therefore potentially results in some errors that a more sophisticated assessment model would avoid. Recent research findings indicate support for a multidimensional model for sexual recidivism risk. This article explores what we already know about the existence of multiple risk dimensions, what those dimensions may be, and which risk factors assess which dimension. Implications for risk assessment procedures are discussed including how risk evaluations using this model might be conducted.

Accuracy of actuarial procedures for assessment of sexual offender recidivism risk may vary across ethnicity

Little is known about whether the accuracy of tools for assessment of sexual offender recidivism risk holds across ethnic minority offenders. I investigated the predictive validity across ethnicity for the RRASOR and the Static-99 actuarial risk assessment procedures in a national cohort of all adult male sex offenders released from prison in Sweden 1993-1997. Subjects ordered out of Sweden upon release from prison were excluded and remaining subjects (N = 1303) divided into three subgroups based on citizenship. Eighty-three percent of the subjects were of Nordic ethnicity, and non-Nordic citizens were either of non-Nordic European (n = 49, hereafter called European) or African Asian descent (n = 128). The two tools were equally accurate among Nordic and European sexual offenders for the prediction of any sexual and any violent nonsexual recidivism. In contrast, neither measure could differentiate African Asian sexual or violent recidivists from nonrecidivists. Compared to European offenders, AfricanAsian offenders had more often sexually victimized a nonrelative or stranger, had higher Static-99 scores, were younger, more often single, and more often homeless. The results require replication, but suggest that the promising predictive validity seen with some risk assessment tools may not generalize across offender ethnicity or migration status. More speculatively, different risk factors or causal chains might be involved in the development or persistence of offending among minority or immigrant sexual abusers.

Stability of the interpretative risk percentages for the RRASOR and Static-99

Current procedures to estimate sex offender recidivism risk typically involve actuarial instruments, either alone or in combination with adjustments based on other considerations. Two of the most commonly employed actuarial instruments for the assessment of sexual recidivism risk are the Rapid Risk Assessment for Sex Offender Recidivism (RRASOR; R. K. Hanson, 1997) and the Static-99 (R. K. Hanson & D. Thornton, 2000). Although multiple studies of the interrater reliability and predictive validity of these instruments have been completed, with very supportive results, there have to date not been any studies of the degree to which the specific risk percentages attached to each instruments score replicate and remain stable despite changes in underlying samples' recidivism base rates. This study, using data from multiple sources, investigated this issue. The findings indicated that the 5-year risk percentages for the RRASOR were replicated and were remarkably stable despite changes in the samples underlying recidivism base rate. The Static-99 5-year risk percentages were mostly replicated and were stable across varying base rates, but to a lesser degree than was found for the RRASOR. Implications of these results are discussed.

The use of actuarials at civil commitment hearings to predict the likelihood of future sexual violence

Some have argued that acturarial methods such as the RRASOR, the MNSOSTR, and the Static-99, can outperform clinical judgments when utilized at a civil commitment hearing to make a prediction. Although actuarial data can be used to identify a group of persons to be considered for possible civil commitment, at present it cannot be used to accurately predict the likelihood of future acts of sexual violence with respect to any specific individual within such a group. For that reason, it might be best to restrict the use of acturial data to the initial screening process, rather than introducing it at the commitment hearing itself.

Differences in the predictive validity of actuarial risk assessments in relation to sex offender type

This study seeks to expand on the previously reported validity of the Static-99, RRASOR, MnSOST-R, and SORAG in predicting sexual recidivism utilizing a regional sample of offenders. The predictive validity of each test was determined utilizing subgroups of the sample based on each offender's known offense history. The effectiveness of each instrument varied depending on offender type. The Static-99 and SORAG were both significantly predictive of sexual, violent, and any recidivism for extra-familial child molesters, and all four tests were predictive of violent or any recidivism in this subgroup. For incest offenders, all four tests were at least moderately predictive of sexual recidivism, whereas the Static-99 and the SORAG were highly predictive of violent or any recidivism. None of the four tests established consistent predictive validity across recidivism categories in regard to rapists or hands-off offenders, however, the Static-99 and the SORAG were significant in terms of sexual recidivism.

Practical considerations in the interview and evaluation of sexual offenders

The evaluation and assessment of sexual offenders is different than any other type of evaluation, and most clinicians are not properly trained to interview this population. This article addresses the clinical and ethical issues particular to the interview, assessment, and evaluation of these types of offenders. It offers both practical information regarding the interview itself, along with an overview of classification systems, paraphilias, and assessment techniques used with this population. In addition, issues related to risk assessment and risk management are also addressed, and an introduction to the use of actuarial risk assessment instruments is provided.

A Multisite Comparison of Actuarial Risk Instruments for Sex Offenders

Four actuarial instruments for the prediction of violent and sexual reoffending (the Violence Risk Appraisal Guide [VRAG], Sex Offender Risk Appraisal Guide [SORAG], Rapid Risk Assessment for Sex Offender Recidivism [RRASOR] and Static-99) were evaluated in 4 samples of sex offenders (N = 396). Although all 4 instruments predicted violent (including sexual) recidivism and recidivism known to be sexually motivated, areas under the receiver operating characteristic (ROC) were consistently higher for the VRAG and the SORAG. The instruments performed better when there were fewer missing items and follow-up time was fixed, with an ROC area up to .84 for the VRAG, for example, under such favorable conditions. Predictive accuracy was higher for child molesters than for rapists, especially for the Static-99 and the RRASOR. Consistent with past research, survival analyses revealed that those offenders high in both psychopathy and sexual deviance were an especially high-risk group.