Section 3: Lecture Content and Teaching Notes
Components of Supervision: Specialized Approaches to Managing Sex Offenders
6 hours, 50 minutes
TOPIC: SEX OFFENDER ASSESSMENT
(60 minutes, including Learning Activity)
|SEX OFFENDER ASSESSMENT BY PROBATION/PAROLE|
Sex offender assessment is an extremely broad topic, and the term can be used by many people to mean many different things. Slide 19 includes a chart that outlines the range of individuals who might be interested in assessing or evaluating a sex offenderfrom the prosecutor to the judge to the treatment provider, and including the probation officer. The tools they use are varied, as are the purposes of their evaluations. We will be focusing on the assessment activities typically undertaken by probation and parole agencies; slide 20 outlines some of the uses of assessment information for those agencies.
An important part of the assessment conducted by probation/parole is the development of a PSI. Clearly, much of the information gathering and interviewing involved in the PSI is evaluative in nature. At every step of the way, the probation/parole officer is attempting to gather information that will provide the best insights into what led the offender to commit the sex offense and how to reduce the likelihood of re-offense. Although some of this information is relevant for risk management (i.e., determining the appropriate conditions of supervision), some of it also provides insights into motivation and modus operandi so that, should an offender be managed in the community, the probation/parole officer can be alerted to warning signs of relapse.
|IDENTIFICATION OF RISK FACTORS|
For purposes of supervision, there are at least two perspectives on risk that are significant for the probation/parole officer. The first is the risk assessment resulting from the application of an actuarial tool. Such tools provide some estimation of the likelihood of future sexual re-offense. As such, actuarial devices provide a backdrop or foundation for community supervision, suggesting more intense levels of control and surveillance for the higher risk offenders.
At the same time, an important part of preparing to make a recommendation for incarceration or community supervision, or to generate a case plan (if community supervision is approved at the PSI phase) is to understand the types of risk factors that each offender presents.
The first type is static risk factors, which are historical or part of the offender's constitution, and don't change. Static risk factors emerge from the empirical studies as being particularly correlated with risk to re-offend sexually and include
Any deviant sexual preference but particularly a sexual preference for children (as measured by plethysmograph* or other objective tool or, if these are not available, then by offender self-report);
*The plethysmograph is a device that measures erectile responses in males to both appropriate and inappropriate stimulus materials.
- Any prior offenses and particularly prior sexual offenses;
- Failure to complete sexual offense treatment;
- Offenders who exhibit sadistic arousal or a high degree of psychopathy;
- Young and never-married offenders;
- Offenders with unrelated victims or male child victims;24
- Child molesters (heterosexual) whose offenses have progressed to genital-to-genital contact.25
Some tools help practitioners use these factors to place offenders in categories according to their risk of re-offending. The probation/parole officer should flag static risk factor information in the file to highlight the dangerousness of the offender and add context to situations that the offender may be in when he is in the community.
Regardless of the formal risk assessments that are used, building an effective supervision strategy demands that the probation/parole officer be aware of potentially risky situations that the offender may put himself into or create. These situations need to be ameliorated immediately. These are referred to as acute dynamic risk factors, as they can change rapidly. They are not necessarily related to long-term recidivism but are important for case management. Acute dynamic risk factors include
- Substance use;
Negative mood (e.g., depression, anxiety);
- Anger, hostility; and
- Access to victims.
Stable dynamic risk factors are more likely to change slowly over time with intervention. Information regarding these items should be gathered routinely as a preparation for completing a PSI or a case plan. Stable dynamic risk factors include
- Intimacy deficitswhether an offender has a stable, satisfying sexual relationship;
- Negative social influenceswhom an offender associates with regularly;
- Attitudeswhether an offender maintains attitudes and values that support sexually abusive behavior;
- Sexual self-regulationwhether an offender has the ability to regulate the emotional and sexual feelings that often precede sex offending; and
- General self-regulationwhether an offender is impulsive and exhibits low self-control.26
You may wonder why denial is not on the list of dynamic factors. The truth is that in terms of statistically validated risk factors, denial is not significant. It does not emerge in the statistical analysis as a distinguishing feature. There may be a number of explanations for this. On one hand, because virtually every offender goes through denial at some point, the mere presence of denial does not serve to distinguish a subgroup of higher risk offenders very well from a statistical point of view. But common sense tells us two things: first, since treatment failure is a significant risk factor for recidivism27 and denial can contribute to treatment failure, denial needs to be taken seriously and addressed as a contributor to risk. Second, if both an offender and his family are in denial, the likelihood that the offender will place himself or be placed in high-risk situations is greatly increased, so denial contributes to recidivism in that way as well.
|LEARNING ACTIVITY: CASE STUDY|
Return to the case study to answer questions 8, and 9, and 10:
- What are your initial thoughts about the risks Mike posed to his victim and to the community? Can you identify any red flags or warning signs in his background that might tell you something about his level of risk? If so, what are they and what do they tell you? Do you ever encounter these red flags and warning signs in your own work? If so, how do they affect the way in which you manage a case?
- Lies and deception play a significant role in the sex offense cycle of many sex offenders. What kinds of things might you have done to verify that Mike was collecting items for his daughters and not residing with young children in his apartment? Have you found in your own work that collateral information is helpful in determining whether or not a sex offender is being honest with you? Based on your own experiences and the information in the case study, who else should the probation officer have contacted to collect more information about this situation?
- Even if Mike were not living with any children, would you consider the collection of children's clothing and toys warning signs or red flags? If so, why? Is this the kind of situation in which the perspective of a sex offender therapist might be helpful? If so, why?
|ASSESSING THE PROBABILITY OF RE-OFFENSE28 |
It is helpful to develop an accurate sense of the statistical probability that a sex offender will re-offend. Although statistics can never predict the future behavior of an individual offender, they can help determine whether an offender is closely matched to a group of offenders for whom we have very good tools for predicting the probability of re-offense as a whole.
Researchers have developed actuarial risk assessment tools that provide very accurate estimates of the probability of failure for the typical criminal offender. Unfortunately, those traditional tools are not very helpful in assessing the likelihood of re-offense of sex offenders. Probation/parole officers often possess a very good sense of the limitations of these tools and recognize that sex offenders often score in a very low-risk category using traditional tools.
The research is beginning to yield a number of tools that are being validated on sex offender populations and can be scored fairly easily and reliably by probation/parole officers. They are far superior to intuition or even to the "clinical judgment" of professionals who have worked in the sex offender supervision field for years and make subjective judgments about the re-offense risk posed by offenders. These actuarial tools are based on statistical models and predict an offender's risk of re-offense by determining how he is similar to other groups of sex offenders for whom re-offense risk is known. These tools consider several characteristics, each of which is given a particular weight. These weights are added in a prescribed manner and yield a risk score. The risk score translates into a risk level or an actual percent likelihood of re-offense. The actuarial approach is the one that insurance companies use to determine the insurance rates for different populations of insured individuals based on the statistical likelihood that they will file a certain type of claim.
The actuarial approach provides objectivity, uniformity, consistency, and equality in the decision making process. It is an easy approach to learn, and its design lends itself to withstand legal challenges. However, as with other prediction methods, this model has limitations. Actuarial assessments primarily focus on general predictors and often ignore case-specific factors, and they typically lack the comprehensiveness common to a thorough clinical assessment approach.
Slide 24 outlines a number of ways and tools that are currently in use to assess the likelihood of recidivism in sex offenders and in other criminal offenders. Two instruments that are very promising for assessing sex offender re-offense risk are the RRASOR29 and the STATIC '99.
The correlation coefficient (or r) is used to measure the extent to which a risk factor is associated with re-offending, and it approximates the percentage difference between those offenders who have a particular risk factor and those who do not. Values under .10 have little practical significance; r values of .10 to .19 are considered relatively small; r values between .20 and .29 are considered moderate; and r values of .30 and above are considered large.30 The r for the RRASOR is .27. The r for the STATIC '99, which incorporates the items that are in the RRASOR, is .33.
Note: Many participants may be averse to statistics and the trainer should provide an explanation of r that meets the audience's knowledge level and experience.
|SEX OFFENDER-SPECIFIC RISK ASSESSMENT INSTRUMENTS |
To give you an idea about how an actuarial risk instrument can be used, let us take a look at the RRASOR and apply it to a case study. The goal of this section is not to make the audience proficient in using the RRASOR or in assessing the re-offense risk of sex offenders. Rather, it will provide an illustration of the actuarial approach using a statistically validated tool that many jurisdictions have found helpful in their efforts to make more informed decisions regarding sex offender supervision. In order to develop the ability to administer any actuarial risk assessment instrument and accurately interpret the tool's results or scores, one must carefully review all of the instrument's scoring materials and practice or be trained under the supervision of a skilled and knowledgeable evaluator.
Note: Trainers should familiarize themselves with the entire RRASOR manual to prepare for teaching this section. Developed by Karl Hanson, of the Solicitor General of Canada's office, the RRASOR and its manualalong with updates on the research can be found at www.sgc.gc.ca/epub/ecorrlist.htm1997.
The RRASOR31 predicts the likelihood of rearrest for a sex offense at 5- and 10-year intervals following a sex offender's release to the community from prison/jail or an inpatient treatment setting. A copy of the instrument is included with the training materials.
The tool contains four items or factors, each of which is assigned a particular score: prior sex offenses, age of the sex offender, relationship of the offender to his victims, and gender of the offender's victims. While a sex offender can receive a score of 0 to 3 on the "prior sex offenses" item, the scores of the other three items can only be 0 or 1. On these three items, offenders score a 1 if they are age 25 or younger at the time of community release, sexually offended against any males, and offended against any nonrelated victims. Overall, a sex offender's score on the RRASOR may range from 0 to 5. Although a score of 6 is theoretically possible on the instrument, no offender in the development samples that were used to construct the test scored a 6. Recidivism rates associated with the different scores on the RRASOR are presented both on the overhead and on the hard copy of the tool that is in your participant materials.
Refer to handout:
A hard copy of the RRASOR is provided with the materials, identical to the slide.
Note: The RRASOR was developed using data from seven different follow-up studies and replicated on an eighth independent sample of sex offenders. The total sample used to generate the test was comprised of 2,592 sex offenders.
|LEARNING ACTIVITY: RRASOR|
One of the benefits of the RRASOR is that it's very easy to use. Unlike some of the other tools on the overhead that require advanced training (like the PCL-R) or information about offenders that can be difficult to obtain (like a plethysmograph score for the SORAG), the RRASOR can be scored by any line probation or parole officer. The RRASOR is not as simple as it looks. All users need to be trained in how to use it and to understand its strengths as well as its limitations.
As a brief, simple example, examine the case of Mr. Jones, a 35-year-old sex offender who was arrested and convicted of molesting a 12-year-old male neighbor. Mr. Jones was also convicted of two prior sex offensesabuse of one boy 15 years ago and another boy 10 years ago.
? Discussion Question: How would we score Mr. Jones?
Based on these facts, Mr. Jones would receive a score of 4 on the RRASOR. He would be assigned one point for each of his two prior sex offense convictions, one point for his selection of a male victim, and one point for his selection of an unrelated victim. The graph on the overhead and in your materials indicates that the sexual re-offense risk of someone who scores 4 on the RRASOR is very highthe specific estimated probability within 5 years is 32.7 percent and within 10 years is 48.6 percent.
|INTERPRETING THE RRASOR SCORE|
The offender's score of 4 means that of all of the sex offenders in the test development sample who had a score of 4, 32.7 percent were estimated to have been arrested for or convicted of committing a new sex offense within 5 years of their community release. Actuarial instruments are, therefore, effective at predicting the re-offense rates of a group of similarly defined offenders. However, these tools cannot necessarily identify whether or not a particular individual in that group will or re-offend. In practice, the common operating assumption is that if a particular offender's score is similar to other offenders who have, for example, a high rate of sexual re-offenses, it is reasonable to consider that offender to be a high risk to re-offend sexually as well.
It is critically important to recognize that the actual percentage re-offense rates noted on any actuarial risk assessment instrument, the RRASOR included, may differ among jurisdictions as a result of different methods of reporting sex offenses, and variations in arrest and prosecution rates. However, on an actuarial tool such as the RRASOR, where each point increase in score is associated with an increase in sexual re-offense rates, the evaluator can be reasonably confident that higher scores correlate with higher sexual re-offense risk.32
? Discussion Question: Are you surprised by the results of the RRASOR in this case? If so, why?
Some researchers have suggested another approach to assessing the risk of re-offensean adjusted actuarial approach. Even those who have proposed such a method suggest that adjustments of actuarial risk assessments should be approached cautiously, knowing that actuarial predictions surpass clinical judgments. The most recent research suggests that adjustments, particularly those based on clinical progress, significantly reduce the accuracy of risk predictions. The CSOM training curriculum on assessment goes into more detail on the theory and practice of adjusted actuarial risk assessment.33
Note: The trainer should anticipate some skepticism regarding the RRASOR because of the fact that there are so few items included on it. If the trainer would like more information regarding the validity (and the degree to which additional items would probably not add much to the predictive validity of the instrument), s/he should refer to the relevant segment of CSOM's training curriculum on assessment.
In recent years, very significant progress has been made in our ability to assess the likelihood of sexual recidivism in sex offenders (the development of the RRASOR is an example of this), and our knowledge in this area continues to develop at an exciting rate. We have identified many static and dynamic risk factors that are statistically correlated with sexual recidivism risk, and researchers have constructed an array of assessment tools that integrate many of these factors in a scientifically valid manner.
These assessment tools significantly enhance our ability to identify subgroups of offenders who pose a higher risk to past and potential victims and the community than others. They also allow us to make more informed management decisions. Like the polygraph, however, none of these tools is a "silver bullet" or a "panacea." Jurisdictions that are working to integrate the tools into their approach to sex offender management must
- Be very clear about why and how they will use the tools;
- Understand fully the strengths and limitations of each (and what they can and cannot do);
- Use more than one, if possible;
- Understand that an offender who scores "low risk" on the tools is not necessarily "low risk" (none of the tools provide all of the answers that we need or want to know about sex offenders); and
- Be willing and able to keep up with the emerging research in this arena.
What probation or parole officers normally think of as assessment may be quite different from what treatment providers, victim advocates, or judges term "assessment." Always try to clarify what the term means and to work collaboratively with your colleagues to define the purposes, the tools, the outcomes, and the issues that are underlying the term "assessment."