The prediction of sexual recidivism is one of the most difficult tasks in forensic psychology since human behavior is dependent upon a variety of different factors. Yet assessing risk of recidivism in sexual offenders is an important societal and legal issue because it guides decisions concerning public safety and adequate treatment planning for sexual offenders. Two actuarial risk assessment instruments (ARAIs), the Stable-2007 and the Acute-2007, encompass risk factors that are able to change slowly (i.e., stable) or quickly (i.e., acute) and are therefore amenable for offender treatment programs to reduce the risk to sexually reoffend. By assessing dynamic risk factors, the Stable-2007 and Acute-2007 predict the probability to reoffend in adult male sexual offenders. This article provides a snapshot of recidivism risk prediction and static and dynamic risk factors, then focuses on the Stable-2007 and the Acute-2007, their development as well as their items, scoring, and validity.
Actuarial Prediction of Recidivism Risk
In order to predict the recidivism risk of sexual offenders, clinicians started to assess offenders based on their subjective experiences after some personal contact with the person concerned. However, Paul E. Meehl demonstrated in 1954 that intuitive judgments, even if they were conducted by experienced clinicians, performed not better than by chance. Meehl revealed that explicitly defined risk factors (e.g., the number of previously committed crimes) that were empirically linked to an increased recidivism risk were more accurate than intuitively established clinical risk factors. He defined the former as the actuarial method of prediction, which has at least two main characteristics: First, there is an explicit rule of combining risk factors (e.g., summing up all scores to one total score), and second, this sum is linked to an empirically derived probability figure (e.g., the probability to reoffend). Structured prediction methods, especially actuarial methods, are state-of-the-art tools in the prediction of recidivism risk in sexual offenders.
Static and Dynamic Risk Factors
Every ARAI consists of a number of risk factors that have to be combined in an explicit manner. These risk factors have different characteristics related to their stability over time. Risk factors are categorized as static or dynamic depending on whether or not they are potentially prone to change. Static risk factors are characteristics of an offender, which are unchangeable, such as the criminal history (e.g., the number of prior offenses) or victim characteristics. Static risk factors provide information about the status risk of an offender compared to the population he belongs to. Due to their clear operationalization, static risk factors can be easily rated based on file information, which leads usually to a high interrater reliability of these ARAIs. Dynamic risk factors are amenable to change and can be further divided into stable or acute dynamic risk factors depending on the rapidity of change. Stable dynamic risk factors last usually longer time periods (e.g., months or years), but they could be potentially modified, for example, by treatment. Examples of stable risk factors are risk-related traits (e.g., lack of concern for others or impulsivity), sexual self-regulation deficits (e.g., sexual preoccupation or deviant sexual interests), or intimacy deficits (e.g., emotional identification with children or loneliness). Acute dynamic risk factors are factors that could rapidly change and are therefore conceptualized as immediate warning signs of an impending risk to commit a crime. Examples of acute risk factors are victim access, substance abuse, or the collapse of the social support system.
The research and proliferation of static risk factors as well as the development of ARAIs consisting of stable and acute dynamic risk factors were strongly influenced and supported by the Risk-Need-Responsivity model of offender rehabilitation, which indicated that an effective intervention has to focus on risk (i.e., the risk potential of the single offender for committing new offenses), need (i.e., consideration of empirically proven criminogenic needs in terms of particular treatment goals), and responsivity (i.e., the use of intervention techniques and treatment programs to which the individual offender’s abilities, learning style, motivation, and strengths respond).
The use of ARAIs can be regarded as an important component of effective offender rehabilitation because it contributes to an accurate measure of the individual level of risk by using static risk factors (i.e., risk principle) and by defining treatment targets in terms of criminogenic needs by using stable and acute dynamic risk factors (i.e., need principle). Currently, the most widely used series of actuarial sexual offender risk assessment instruments are the Static-99, which consists of 10 static, historical, and biographical risk factors; the Stable-2007, which consists of 13 stable dynamic risk factors; and the Acute-2007, which consists of seven acute dynamic risk factors.
The Development of the Stable-2007 and the Acute-2007
These two instruments were developed to assess dynamic risk factors in order to predict the probability to reoffend in adult male sexual offenders. There have been three empirical steps in the development of the current versions of the Stable-2007 and the Acute-2007.
The predecessor of both instruments, the Sex Offender Risk Assessment Rating, was developed by Karl Hanson and Andrew Harris within the Dynamic Predictors Project that spanned from 1996 to 2004. In this project, the researchers compared 208 sexual offenders who reoffended during their time on community supervision with 201 offenders who did not reoffend during that time period. When investigating the differences between these groups, they found two different types of dynamic risk factors: stable and rapidly changing factors. Therefore, the Sex Offender Risk Assessment Rating implied five stable and four acute dynamic risk factors.
In the second step, Hanson and his colleagues revised the Sex Offender Risk Assessment Rating due to conceptual and clinical concerns by separating stable and acute items into two distinct measures, the Stable-2000 and the Acute-2000. Additionally, the items of the Stable-2000 were further embedded in a more psychologically sound theory. A manual provided clear operationalizations and definitions of all items as well as explicit scoring details. The Stable-2000 comprised six risk areas: significant social influences, intimacy deficits, sexual self-regulation, general self- regulation, cooperation with supervision, and attitudes supportive of sexual offending. The final score results from the highest score of an item within each risk area. The Acute-2000 contained 8 items: seven general risk factors and a unique risk factor which should capture risk-relevant information only applicable in the individual case.
In the third step, Hanson and colleagues started a study in 2007 by using a prospective-longitudinal research design, the Dynamic Supervision Project. Parole and probation officers in Canada, Alaska, and Iowa were trained in the administration of the Static-99, the Stable-2000, and the Acute-2000. Across 16 jurisdictions, 997 sexual offenders were assessed; after a median of 41-month follow-up time period, the predictive accuracy of the three instruments was investigated. Based on these results, the Stable-2000 and the Acute-2000 were revised to their current versions (see Tables 1 and 2). Due to a relatively low predictive accuracy of the Stable-2000, the ratings of 3 items were adjusted. The risk area scores of the Stable-2000 became less important, and all items were added up to yield the final risk score. Additionally, the 3 items concerning attitudes supportive of sexual offending were dropped because they did not significantly predict sexual reoffending. With regard to the Acute-2000, the individual risk factor was dropped, and the researchers examined two different dimensions within the instrument. The first dimension was especially relevant for the prediction of sexual and violent recidivism and consisted of 4 items, while the second dimension represented a general criminality factor which included all seven risk factors of the instrument.
Stable-2007
The current Stable-2007 version is an ARAI for sexual offenders comprising 13 stable dynamic items allocated to five risk domains. These domains and items are presented in Table 1.
The use of the Stable-2007 requires a thorough review of the files of the offender and an interview with him. Each risk factor has to be scored by considering whether or not this particular factor would play a risk-relevant role in the next 12 months assuming that the offender would have the opportunity to reoffend (i.e., that he is actually released and on risk). Each item is scored on a 3-point rating scale, with 0 for no problem, 1 for some concern/slight problem, and 2 for present/ definite concern. Total scores are obtained by summing up all item ratings with higher scores indicating higher risk to recidivate. Total scores ranging from 0 to 26 are transformed into three risk categories: low, medium, and high. Additionally, these risk categories may be combined with risk categories of the Static-99 to obtain combined risk-need categories.
Measurement with Stable-2007 is very precise as shown by interrater reliability indices of .90 or higher (values of .75 and above typically are considered excellent according to conventional standards). Furthermore, the Stable-2007 is a strong predictor of sexual recidivism which is demonstrated by moderate to large effect sizes for the prediction of sexual recidivism in male adult sexual offenders in various studies from different countries and jurisdictions. Besides the Dynamic Supervision Project-based studies, there were two independent cross-validation studies from Austria which revealed significant effect sizes for the prediction of sexual recidivism with Area Under the Curve (AUC) values ranging between .67 and .71 for male sexual offenders released from the prison system as well as for forensic patients. Furthermore, the S table-2007 has been shown to have incremental validity beyond static risk factors in different Canadian and Austrian samples.
Acute-2007
The current version of the Acute-2007 encompasses seven acute dynamic risk factors that are rated in four different categories from 0 for no problem, 1 for maybe/some problem, 2 yes, definite problem, and to 3 for intervene now. The last category intervene now indicates situations in which the risk of reoffending may be sufficiently high to initiate immediate preventive interventions.
Table 2 lists the items of the Acute-2007. For the prediction of sexual recidivism, items 1–4 are summed; for the prediction of general recidivism, all items are summed for the final score rating. Both final scores aimed to predict recidivism within the next month and can be transformed into three risk categories: low, medium, and high. Additionally, these risk categories can be combined with the combined Static-99 and S table-2007 risk-need categories to inform risk management decision-making processes and to further improve the prediction of recidivism. There has only been one comprehensive empirical investigation published about the reliability and validity of the Acute-2007, which is based on the Dynamic Supervision Project database from Hanson, Harris, Terri-Lynne Scott, and Leslie Helmus in 2007. First, the findings support the predictive and incremental validity of acute risk factors. The results showed a high effect size for the sexual recidivism subscale and a moderate effect size for the general recidivism subscale in predicting recidivism for sexual offenders under community supervision. Second, a more in-depth analysis of the items included in the Acute-2007 indicated that the instrument is actually assessing dynamic, changeable risk-relevant propensities for offending.
References:
- Eher, R., Matthes, A., Schilling, F., Haubner-MacLean, T., & Rettenberger, M. (2012). Dynamic risk assessment in sexual offenders using STABLE-2000 and the STABLE-2007: An investigation of predictive and incremental validity. Sexual Abuse: A Journal of Research and Treatment, 24, 5–28. doi:10.1177/1079063211403164
- Hanson, R. K., Harris, A. J. R, Scott, T. L., & Helmus, L. (2007). Assessing the risk of sexual offenders on community supervision: The Dynamic Supervision Project (Corrections research user report 2007-05). Ottawa, Canada: Public Safety Canada. Retrieved from https://www.publicsafety.gc.ca/cnt/rsrcs/pblctns/ ssssng-rsk-sxl-ffndrs/ssssng-rsk-sxl-ffndrs-eng.pdf
- Hanson, R. K., Helmus, L.-M., & Harris, A. J. R. (2015). Assessing the risk and needs of supervised sexual offenders: A prospective study using STABLE-2007, Static-99R, and Static-2002R. Criminal Justice and Behavior, 42, 1205–1224. doi:10.1177/0093854815602094
- Rettenberger, M., & Craig, L. A. (2016). Actuarial risk assessment of sexual offenders. In L. A. Craig & M. Rettenberger (Eds.), The Wiley handbook on the theories, assessment and treatment of sexual offending—Volume II: Assessment (pp. 609–642). Chichester, UK: Wiley.