WED, Oct 6 \ 14.30-16.30 \ Risk Assessment, I

006
A THREE-FACTOR CONCEPTUAL MODEL TO COMBINE STATIC AND DYNAMIC RISK WITH TREATMENT AND SUPERVISION STRATEGIES IN THE MANAGEMENT OF SEX OFFENDERS

Howard E. Barbaree, Ph.D, Calvin M. Langton, Ph.D.
Law and Mental Health Program, Centre for Addiction and Mental Health and the University of Toronto, Canada
Howard_Barbaree@camh.net

During the past 10-12 years, considerable progress has been made in the field of risk prediction with sex offenders. Research on static or historical factors such as an offender's developmental history have led to the promulgation of empirically validated actuarial risk instruments that are demonstrably predictive of recidivism. More recently, a distinction has been made between static and dynamic risk factors. In distinction to the static historical factors, dynamic risk factors such as mood and fantasy may change over time. Recent research with sex offenders has identified a number of dynamic factors that seem to change before an offender recidivates. Dynamic factors could be used to indicate when an offender is at greater risk to reoffend. Being thus alerted, a case manager could intervene to prevent recidivism. While such case management strategies are often used with sex offenders, current treatment models do not recognize the contribution of such risk management interventions. A three-factor model is presented to conceptualize the critical components of risk management to prevent recidivism in sex offenders. The first factor utilizes actuarial assessment to stratify offenders into levels of risk (low, moderate, high). Subsequent risk management resources are divided among the strata such that greater resources are devoted to higher risk offenders. The second factor, referred to as supervision, combines dynamic risk assessment with case management interventions designed to prevent recidivism (e.g., parole suspension, increased restrictions, etc.). The final factor is treatment. Well known cognitive-behavioral treatment strategies are combined to support case management and supervision. Data are presented to illustrate how the factors combine to reduce recidivism in sex offenders.

007
DETERMINING RISK: COMBINING ACTUARIAL AND CLINICAL RISK ASSESSMENT

Don Grubin, MD, MRCPsych.
Professor of Forensic Psychiatry, (Hon) Consultant Forensic Psychiatrist, University of Newcastle, Newcastle upon Tyne, England
don.grubin@ncl.ac.uk

Although actuarial assessment of reoffending is a good indicator of long term risk, on its own it is of limited utility in the day to day supervision of offenders. Combined with clinical and other types of assessment of dynamic types of risk factor, however, it can provide a useful baseline from which to make management decisions. In this presentation a model will be described in which static and dynamic risk factors are combined in a systematic manner both to generate management protocols, and to assist probation officers and clinicians in deciding when and how to intervene. Developed in a probation context, this model of risk assessment is currently being piloted in an English probation area.

008
PHALLOMETRIC TESTING WITH SEXUAL OFFENDERS AGAINST FEMALE VICTIMS: AN EXAMINATION OF RELIABILITY AND VALIDITY ISSUES

Yolanda Fernandez, Ph.D.
Correctional Services of Canada fernandezym@csc-scc.gc.ca

The present series of studies explored issues related to the reliability (internal consistency and test-retest reliability) and criterion validity of phallometric testing with sexual offenders (incest offenders, extrafamilial child molesters, and rapists). Three assessment sets were evaluated: (1) An Age-Gender set which presented slides of adults and children; (2) A Female Sexual Violence set which presented audio descriptions of consenting and forced sex between adults; and (3) A Child Sexual Violence set which presented audio descriptions of the sexual molestation of children by adults. A total of 280 incest offenders, 138 extrafamilial child molesters, and 139 rapists were included in the different analyses.
Study 1 demonstrated that the internal consistency of the three assessment sets was satisfactory. Unfortunately, the test-retest reliability of two of the assessment sets (Age-Gender and Female Sexual Violence) was less than acceptable although further analyses suggested that habituation and practice effects (faking) did not influence the scores between testing occasions.
In terms of criterion validity using the Age-Gender set, it was found that the extrafamilial offenders were more deviant than the incest offenders and they demonstrated a sexual preference for children that was not apparent among the incest offenders. In response to the Child Sexual Violence set the incest offenders and extrafamilial child molesters demonstrated similar response levels with the Pedophile Index indicating a preference for the child stimuli among both groups. However, responses to both the Age-Gender and Child Sexual Violence sets did not accurately
identify group membership. Within the incest offender group subjects appeared more deviant when assessed using the audiotaped stimuli than when using the slide stimuli.
Surprisingly the rapists did not appear to respond more deviantly than did either the incest offenders or the extrafamilial child molesters to the Female Sexual Violence set and none of the three groups demonstrated a sexual preference for rape or violent behaviors over consenting sexual acts. In addition, the Female Sexual Violence Assessment did not accurately predict group membership. The overall results are discussed in terms of their relevance for research, treatment and theories of sexual offending.

009
RELATIVELY SIMILAR VIOLENCE PREDICTORS: WHY EVERYTHING WORKS

Douglas P. Boer, Ph.D.,
Correctional Service of Canada
boerinc@telus.net

The main approaches to risk assessment of sexual offenders are compared and contrasted. While each approach has its merits and problems, one common denominator is apparent: all assessment instruments designed for sexual offender risk assessment predict sexual offending to some significant degree. In fact, some instruments not designed for this purpose also predict sexual offending. The simplest reason for this is the apparent overlap between instruments in terms of the variables used. That is, the variables that comprise the majority of these instruments are relatively similar in nature. The variables that comprise a number of actuarial and structured clinical guideline tests are examined to support this claim. As a result, the actuarial-structured clinical guideline debate is seen as a moot discussion as these types of tests are more similar than the authors may admit. Finally, given the similarities, it ought not to be a surprise that neither approach has convincingly proven itself superior statistically or clinically to the other.