Forensic psychology: Part 4: Chapter 9: Assessing offender populations

by Sharon M. Williams, Ph.D., C. Psych ^{Footnote 1} 

Objectives
1) To identify guidelines for referral agents.
2) To identify key theoretical and practical issues in the assessment of sex offenders.
3) To examine professional qualifications of practitioners.
4) To provide a bibliography of further selected readings.

Prologue
The evaluation of sex offenders has become both increasingly complex and highly visible. The sex offender population continues to increase at a higher rate than the general population, and public scrutiny has become more intense.

As of December 1992 (NHQ, 1993), the federal sex offender population totalled 3,700, representing more than 17% of the total offender population. This represents a growth of 28% over a four year period. Despite generous resourcing, demand for assessment and treatment consistently outstrips available resources. Resources have also been directed toward research into sex offender assessment, especially in the area of risk prediction and management.

Guidelines for referral agents
Assessment of sex offenders has particular relevance to the offender's correctional program, as well as treatment and risk management in the community. Referral agents should become familiar with the Criminal Code as it relates to sex offenders. The new Code includes 64 sexual offences (see Appendix A), ranging from the obscure (seduce female passenger on vessel) to the more common — rape and sexual assault.

Unfortunately, on some occasions, plea bargaining may reduce a sexual assault to an offence such as break and enter, trespassing, or assault causing bodily harm. In the case of a murder conviction, the most serious offence is noted in the conviction, although a sexual offence(s) may also have occurred.

Therefore, familiarity with the details of the case is essential when making a referral for appropriate evaluation.

The following would be appropriate questions when making an assessment for referral:

• What are the offender's current deficits/ problem areas?
• 'What program(s) would best suit the offender's needs?
• Where and when should this program take place?
• 'What is the best estimate of the offender's motivation and treatment potential?
• What intervention strategies would be useful adjuncts to specialized services (i.e., substance abuse, cognitive skills), and when should these take place?

Once an intervention strategy is carried out, appropriate questions would be:

• Has risk been reduced, and to what degree?
• What is the offender's current level of risk?
• Can the offender's risk to reoffend be further reduced/managed, and in what manner can this be done?
• What is the offender's crime cycle, high-risk situations and how are they to be managed?
• What is the offender's relapse prevention plan?
• How well can this plan be put in place?
• What is the best method of monitoring this individual in the community?
• If a lapse occurs, what is the best method of dealing with it?
• What are the warning signals that a relapse is imminent, and how should this be managed?

i) Phallometrics
The assessment of sex offenders is a relatively new field with a number of emerging trends. Psychometric and phallometric evaluation, file review, standardized interviews, and behavioural assessments are being developed and refined in a number of centres (Barbaree et al., 1994; Knight, Prentky & Cerce, 1994) and should be considered emerging trends rather than definitive or proscriptive procedures.

Measuring sexual arousal can add precision to the assessment of sexually deviant interests. Phallometric evaluation has been intensely scrutinized over the past three decades. In most cases, evaluating penile tumescence to various stimuli has involved placing a mercury-in-rubber strain gauge approximately midway along the penis. The circumference of the penis increases with arousal, and these changes can be sampled virtually on a continuous basis by computer. The use of a volumetric device provides a more precise measurement, but one which is more difficult to obtain.

Testing must occur in a quiet area where the offender is physically separated from the technician. The offender will be partially unclothed and will need to have a towel placed across his lap. An intercom system should be in place to allow efficient communication between the offender and the technician. It is recommended that a closed circuit camera be used to monitor the offender in order to reduce voluntary distortion of results.

Prior to assessment, the offender should be familiarized with the equipment, laboratory, and procedures used in phallometric evaluation. This will reduce anxiety and permit informed consent from the offender. The offender should be permitted to terminate the assessment if he is opposed to any of its components.

ii) Age-gender and sexual violence assessments
For the age-gender evaluation, the offender will be exposed to a variety of slides, half of which involve males, and the other half females. The ages of the people in the slides should fit into four main categories: prepubescent, pubescent, adult, and neutral (Malcolm, Andrews and Quinsey, 1993). A number of rules exist with respect to warm-ups, slide duration, randomization of slides, and return to baseline. These issues are complex and cannot be appropriately dealt with in this context (Harris, Rice, Quinsey, Chaplin and Earls, 1992). In order to deal with the question of reliability, the offender should be exposed to the age-gender assessment on two occasions, preferably at least one day apart.

Violence is most commonly assessed through the use of audiotaped stimuli which depict neutral, consensual and non-consensual sexual interactions. These can involve children or adults of either sex. The tapes used in the Ontario Region were developed at the Mental Health Centre (Penetanguishene) and have been described in detail by Lalumière and Quinsey (1994). It has been suggested that maximally violent stimuli are the most effective in discriminating between samples of offenders and "normals."

Scoring of results has taken a variety of formats. However, the best procedure appears to involve conversion of raw scores to Z scores, which can then be manipulated statistically. Phallometric evaluation possesses discriminant and predictive validity. In other words, this form of evaluation provides valuable incremental information which can add to information already obtained through interviews, file review, and psychometric data. It is also one component of the Statistical Prediction of Violent Recidivism by Sex Offenders (Harris, Rice & Quinsey, 1993; Quinsey, Rice, Harris & Lalumière [in press]) which can then improve risk prediction.

Unfortunately, information obtained through phallometric evaluation is not always easily interpretable. Offenders may attempt to distort results by obvious means such as dosing their eyes, lifting the penonometer with a finger or pencil, using pain (such as a thumb tack) to reduce response, trying to imagine alternate stimuli, or using a "pumping" motion, a contraction of the pubococcygeal muscles to enhance responding. These distortions can be reduced by observation through dosed circuit monitors, and by careful examination of the data. However, voluntary distortion of results remains a problem which is being examined by researchers.

Although some pedophiles show a "normal" response (more response to adults than children) and some sexual aggressors who offend against adults show a similar preference for consensual stimuli over coercive, when deviant preference is found, regardless of absolute arousal, there is evidence that risk is increased. Thus, a "normal" pattern of responding can be considered low risk, while a lack of differentiation in responding can be considered moderate risk (Barbaree personal communication, 1994) and deviant responding is generally considered a high risk factor. In summary, arousal assessment provides incremental information which impacts on both risk assessment and intervention strategies.

iii) Typology
A second area of interest in the assessment of sex offenders is typology. The development of increasingly precise methods of classifying offenders, and relating these factors to risk is underway (Knight et al., 1994). File review and precise clinical interviews can provide useful information in determining the degree of risk which the offender may demonstrate, and which intervention strategies may be most suited to the offender's particular needs. Historical factors (such as early behaviour problems) are static risk predictors, but there are a number of more dynamic factors, including phallometric response, sexual attitudes, and attitude towards crime, which may be responsive to intervention techniques.

Although sub-classification of offenders is a detailed task, determining both motive and skill level can result in the sub-division of sexual assaulters. Categories such as pervasively angry, non-sadistic, sadistic, opportunistic, and vindictive can be established. Within each group, further subcategories such as skill level are possible. Barbaree et al. (1994) found only one pervasively angry aggressor and, therefore, dropped this group from analysis.

Pedophiles can be sub-divided by age of victim, frequency of assaults, intrusiveness, and use of a weapon. Incest offenders are often excluded from studies as they are most similar to non-incarcerates and have an exceedingly low likelihood of recidivating (Khanna, Malcolm, Brown & Williams,1989). Prentky (1994) has also found that gender of victim does not account for a significant portion of the variance once frequency of sexual offending and other more prominent variables are controlled such as: amount of time spent with children, strength of sexual interest, impulsivity, antisocial history, paraphilias, substance abuse, and social competence.

iv) Psychometrics
At Millhaven Institution (Ontario), an assessment battery has been developed to examine five psychometric areas: intelligence, sexual knowledge and attitudes, sexual history, risk scales, and social desirability bias (see Appendix B for information on Millhaven's Sex Offender Intake Services). A number of other areas can be evaluated such as hostility, hostility towards women, empathy, sexual issues (gender identity, repression) assertiveness, and sexual anxiety. Walbek (1993) reports that the Multiphasic Sex Inventory developed by Nichols & Molinder (1984, 1992) is useful in considering changes during treatment and at post-treatment.

v) File review and structured interview
Risk evaluations, such as those developed by Hare (1991) and Harris, Rice & Quinsey (1993), are dependent on accurate gathering of information. Although only a small proportion of sex offenders have been identified as high risk on the PCLR (Barbaree et al., 1994), that information, along with phallometric and psychometric data, can help in assigning a risk level. Other risk instruments such as the LSI (Andrews, 1983) and SIR scale (NufBeld, 1989) are also useful.

In a structured interview, areas which should be covered include family history, early development, educational history, relationship with peers, first sexual interaction, pattern of sexual interest (thoughts, fantasies), history of criminal involvement (juvenile and adult), employment history, common-law relationships, marriages, and other sexual involvements. A detailed examination of the sexually deviant behaviour and its precursors will be of primary importance. A history of releases and reasons, if any, for their termination should be considered. Substance abuse history and positive leisure activities should also be examined. Finally, motivation for treatment and attitude towards both crime and victim will be essential when evaluating suitability for intervention.

vi) Behavioural assessment
Although time-consuming, a videotaped assessment of the offender's social skills can be a valuable supplement to the assessment process. While it might be possible to voluntarily distort responses to psychometric instruments, it is considerably more difficult to overtly demonstrate a complex skill without having the requisite behavioural elements. Behavioural assessment of conversational skill, assertiveness and empathy have been developed for use with offenders at the Regional Treatment Centre (Ontario) and the Regional Psychiatric Centre (Prairies).

Briefly, the conversational skill assessment involves a five-minute conversation with male and female confederates. These are scored either by the confederate, or by a "blind" rater on two dimensions: anxiety and skill. The assertiveness assessment involves twenty short scenarios involving both male and female confederates in "street" and prison situations. The empathy assessment (Williams and Khanna, 1987, 1990) involves a female confederate who has a serious problem. The offender must identify the problem and show both understanding and feelings for the confederate. This measure can discriminate between offenders and non-offenders, and is responsive to changes resulting from treatment.

In summary, thorough assessment which involves file review, structured interviews, psychometric and phallometric evaluation will help to identify the factors which influence the risk to reoffend sexually. Identification of these factors, some of which will be static, others dynamic, will impact on the decision to offer treatment, as well as the intensity, duration, and location of further programs.

vii)Treatment/intervention strategies
Assessment has identified several major difficulties faced by many sex offenders. These may include:

• cognitive distortions surrounding sexual offending;
• an inability or lack of motivation to interfere with the crime cycle;
• deficits in social skills, including anger management and empathy, which may affect the development of intimate relationships;
• deviant arousal; and
• substance abuse.

As this document is primarily aimed at examining issues related to the assessment process, it will not be possible to offer more than a brief review of intervention approaches.

Most treatment programs in Canada claim to be based on a cognitive-behavioural model (NHQ, 1993), which focuses on altering the emotions, thoughts, and behaviours which influence offending. Group format is frequently cited as the primary method of treatment delivery as it is both cost effective and useful as a means of peer confrontation of attitudes/cognitions. Most programs, however, also use individualized sessions which can address deviant arousal, lead to improved understanding of the offender's idiosyncratic crime cycle, address particular social inadequacies, and can allow for thorough discussion of the offender's own victimization and its emotional and attitudinal sequelae.

Programs tend to target denial, minimization, and rationalization early in therapy. Although a variety of procedures can be used, most involve a group component where the offender publicly discloses his version of the offence. There is some attempt to dovetail the offender's version with the official police report and with the victim impact statement. Programs vary in the vigour with which they pursue concordance, and some rely on individual therapy to accomplish this process.

Programs which are low intensity and aimed at low risk/low needs offenders use a relapse prevention model (Laws, 1989, Pithers, 1990), and examine crime cycle, high-risk situations, and methods of escaping or avoiding them. Victim empathy is also addressed. Programs for low-risk offenders are often open-ended and of short duration (8-10 weeks).

For offenders who have more extensive treatment requirements and who are at higher risk to reoffend, programs tend to include a module on social skills enhancement, attitudes towards sexuality, knowledge about sex, and often target deviant sexual arousal. Victim empathy and relapse prevention are also treatment components for the higher needs/risk group. Other idiosyncratic needs such as anxiety management and dealing with victimization may be addressed individually. Program length tends to vary from 5 months to 8 months. Programs may also be physically located in medium-security institutions or in residential locations such as the RPCs or RTCs. With respect to intensity, programs last approximately 15 hours per week. After program completion, relapse prevention is often offered at minimum security prisons as well as in the community. This broad spectrum approach should effectively meet the needs of most sex offenders.

There are occasions when sex offenders cannot control their sexual urges despite good motivation and involvement in comprehensive programs. Use of chemical "castrators" such as MPA (Medoxyprogesterone Acetate) and CPA (Cyproterone Acetate) (Cooper, 1986), or serotonergic reuptake inhibitors such as Lupron (Federoff, 1994), can reduce deviant sexual fantasies and deviant behaviours. While there are still problems in maintaining compliance with MPA and CPA regimes, Federoff (1994) found Lupron to be less problematic.

The literature is polarized on the question of whether treatment for sex offenders is effective. Quinsey, Harris, Rice and Lalumière (1993) argue that there is insufficient methodologically sound evidence in favour of treatment efficacy. Marshall (1993) and Marshall and Pithers (1994) argue equally forcefully that, methodological issues aside, treatment has been shown to be effective in reducing cost to society (victimization, trials, incarceration). A large scale study by Marques (1995) is underway in California. Due to a number of fortuitous factors such as inadequate funding, this study involves random assignment to treatment and control groups. Some preliminary findings have been presented, but the small numbers released for an adequate time period make it difficult to interpret the results. However, it is evident that the untreated volunteers have a high rate of return and treatment does appear to have some impact.

Providing sex offender assessment and treatment calls for a variety of qualifications, depending on the roles played during this process.

Phallometric assessment requires a technician who has been trained at community college, or who has a B.A. Honours and additional supervision in the use of phallometric evaluation. Interpretation of phallometric results has traditionally been carried out by a psychologist with sufficient expertise to evaluate the often subtle changes in arousal patterns.

Psychometric evaluation and file review can be carried out by employees supervised by a psychologist. Interpretation of test results again requires a psychologist. Under the CCRA, the psychologist must be registered or supervised by a registered psychologist.

Draft national standards for the provision of services to sex offenders have been produced by a national committee (National Committee on Sex Offender Strategy, 1995). These standards are currently undergoing revision, but are expected to be promulgated in 1995.

Sex offending is a complex issue due to the heterogeneity of the offenders and the newly emerging trends in the literature. Assessment is an essential first stage in developing an appropriate intervention plan and in evaluating treatment outcome. In addition, the risk posed to the community by the offender is of paramount importance, and must be one of the more salient goals of the assessment process.

A combination of information sources will yield the most accurate snapshot of the offender's characteristics. Psychometric data, behavioural assessment, interview, file review and phallometric evaluation can all yield useful information.

Efficacy of treatment programs is still under review, but the most promising programs should match needs and risk to program content. Low risk and low needs offenders should receive short term, low intensity programs. At the other end of the continuum, high risk/high needs offenders should receive more intense, comprehensive, longer term programs and community follow-up with more intensive supervision. The use of detention to warrant expiry date can interfere with the process of follow-up and should be carefully evaluated.

Chapter 9 Appendix A, B

Andrews, D.A., Kiessling, O.J. & S. Kommos. (1983). The Level of Supervision Inventory (LSI-6) Interview and Scoring Guide. Toronto, Ontario: Ministry of Correctional Services.

Barbaree, H.E., Seto, H.C., Serin, R.C., Amos, N.L. & D.L. Preston. (1994). "Comparison between sexual and nonsexual rapist subtypes," Criminal Justice and Behavior, 21 (1), 95-114.

Cooper, A.J. (1986). "Progestogen in the treatment of male sex offenders: A review," Canadian Journal of Psychiatry, 31, 73-79.

Federoff, J.P. (1994). Treatment of Sex Offenders with Serotenergic Medications. Clarke Conference on Assessment and Management of Risk in the Sex Offender, Toronto.

Hare, R. (1991). Manual for the Revised Psychopathy Checklist. Toronto: Multihealth Systems.

Harris, G.T., Rice, M.E. & V.L. Quinsey. (1993). "Violent recidivism of mentally disordered offenders: The development of a statistical prediction instrument," Criminal Justice and Behavior, 20, 315-335.

Harris, G.T., Rice, M.E., Quinsey, VIL., Chaplin, T.C. & C. Earls. (1992). "Maximizing the discriminant validity of phallometric assessment data," Psychological Assessment, 4 (4), 502-511.

Khanna, A., Brown, P., Malcolm, P.B. & S.M. Williams. (1989). Outcome Data on Sex Offenders Assessed and Treated at the Regional Treatment Centre (Ontario). First Annual Research Conference, Ottawa.

Knight, R.A., Prentky, R.A. & D.D. Cerce. (1994). "The development, reliability and validity of an inventory for the multidimensional assessment of sex and aggression," Criminal Justice and Behavior, 21 (1), 72-94.

Lalumière, M.L. & V.L. Quinsey. (1994). "The discriminability of rapists from non-sex offenders using phallometric measures: A metaanalysis," Criminal Justice and Behavior, 27 (1), 150-175.

Laws, D.R. (1989). Relapse Prevention with Sex Offenders. N.Y: Guildford.

Malcolm, P.B., Andrews, D.A. & V.L. Quinsey. (1993). "Discriminant and predictive validity of phallometrically measured age and gender preference," Journal of Interpersonal Violence, 8(4), 486-501.

Marques, J. (1995). Does Treatment Work? Findings and Lessons from California's Outcome Study. Paper presented at Towards a National Strategy - A Conference on Intervention with Sex Offenders, Toronto, 1995.

Marshall, W.L. (1993). "The treatment of sex offenders: What does the outcome data tell us? A reply to Quinsey, Harris, Rice and Lalumière," Journal of Interpersonal Violence, 8(4), 524-530.

Marshall, W.L. & W.D. Pithers. (1994). "A reconsideration of treatment outcome with sex offenders," Criminal Justice and Behavior, 21 (1),10-27.

National Committee on Sex Offender Strategy. (1995). National Sex Offender Strategy. Draft.

National Headquarters. (1993). Treatment Capacity for Sex Offenders. Ottawa: Correctional Service of Canada.

Nichols, H.R. & I. Molinder. (1984). Multiphasic Sex Inventory Manual, Tacoma, Washington: Nichols & Molinder.

Nuffield, J. (1989). "The "SIR" Scale: Some reflections on its applications," Forum on Corrections Research, 1 (2), 19-22.

Pithers, W.D. (1990). "Relapse prevention with sexual aggressors. A method for maintaining therapeutic change and enhancing external supervision," in W.L. Marshall, D.R. Laws and H.E. Barbaree (Eds.), The Handbook of Sexual Assault, Theories and Treatment of the Offender (pp. 343-361). N.Y.: Plenum.

Prentky, R. (1994). A Taxonomy for Child Molesters and the Assessment of Risk. Clarke Conference on Assessment and Management of Risk in the Sex Offender, Toronto.

Quinsey, V.L., Harris, G.T., Rice, M.E. & M.L. Lalumière. (1993). "Assessing treatment efficacy in outcome studies of sex offenders," Journal of Interpersonal Violence, 8(4), 512-523.

Quinsey, V.L., Rice, M.E., Harris, G.T. & M.L. Lalumière. (In press). "Predicting sex offences," in J. Campbell ( Eds.), Assessing Dangerousness.

Walbeck, N.H., Haroldson, P. & R. Johnson. (1993). Multiphasic Sex Inventory Scores with Sexual Offenders against Children. Presented at the Association for Treatment of Sexual Abusers, Boston.

Williams, S.M. & A. Khanna. (1987). Empathy Training for Incarcerated Sex Offenders. Paper delivered at Ontario Psychological Association, Toronto.

Williams, S.M. & A. Khanna. (1990). "Empathy training for sex offenders," Proceedings of the Third Symposium on Violence and Aggression. Published by the University of Saskatchewan and the Regional Psychiatric Centre (Prairies).