Mental health practitioners are often asked to provide an opinion on the likelihood that an individual will become aggressive or violent, as the criminal justice system aspires not just to punish past criminal behavior but also to reduce future criminal behavior. Formalized criminal risk assessments are one approach to determining and assessing relevant risk factors in order to estimate the probability of reoffending and developing plans to minimize risk. Risk factors refer to personal and environmental characteristics that are shown to be associated with an increased risk of future criminal behavior. Static risk factors are those that do not change over time, generally historical information or other factors unchanged by intervention, whereas dynamic risk factors are those that could possibly change, and they are generally considered more amenable to intervention.
Criminal risk assessment strategies are beneficial in a number of contexts including, but not limited to, informing courts, and assessing supervision and treatment needs. More specifically, the issue of criminal risk assessment commonly arises to aid the courts in making legal decisions such as whether to grant bail or parole, sentencing, and civil commitment. A high-risk individual, for example, may have different supervision requirements with regard to probation, parole, or possible civil commitment than someone deemed low risk of reoffending. The emphasis placed on treatment involvement may also vary depending on presenting risk factors and level of risk.
Given the vast implications and significant impact, not only on the community but also on the individuals being assessed, accurate and efficient prediction remains the ultimate objective. Since 2005, there has been a significant increase in the amount of research dedicated to understanding theories of violence, predicting violence, and improving risk assessment procedures. Many instruments, some of which are discussed in this article, have been created with the goal of improving accuracy in the process of violence and/or sexual risk prediction.
Static and Dynamic Risk Factors
Researchers and evaluators conceptualize risk factors in two groups: static and dynamic factors. Static, or fixed, risk factors include historical information that will remain constant across time. Dynamic, or changeable, risk factors are those that are modifiable with interventions. It is important for practitioners to keep both types of risk factors in mind when conducting a risk assessment (see Table 1).
Static Risk Factors
One static risk factor that is a strong predictor of future risk is a history of violence. Research suggests other relevant static risk factors include male gender, younger age, and an extensive criminal record. Some researchers argue that static risk factors are more useful for identifying long-term risk; however, static risk factors are often deficient in being able to predict or indicate when violence or aggression may occur. Furthermore, many researchers emphasize that static risk factors are not generally useful in determining potential change in risk over time.
Dynamic Risk Factors
It has been suggested that dynamic risk factors are vital to criminal risk assessment because they can assist in identifying those individuals who are the best candidates for intervention. E xamples of dynamic risk factors include the following: interpersonal relationships at home, school, or work; level of association with delinquent peers or associates; and antisocial attitudes and/or personality.
Dynamic risk factors can further be differentiated into stable dynamic risk factors and acute dynamic risk factors. Stable dynamic factors are those risk factors that are generally enduring characteristics (e.g., problems related to substance abuse, antisocial personality). These are likely to remain unchanged for months or even years. It has been argued that interventions with the goal of creating lasting improvements should target stable dynamic risk factors. Acute dynamic factors are those risk factors that are more rapidly changing and might immediately precipitate the perpetration of an offense (e.g., intoxication, negative mood). These may last days, hours, or even minutes. While acute dynamic risk factors are specifically related to the timing of offending behavior, they may have little connection with long-term risk likelihood.
Some dynamic risk factors that show a negative relationship with reoffending are categorized as protective factors. In other words, the presence of these factors appears to reduce the risk of future violent behavior. These can be personal characteristics or environmental/situational characteristics. Some protective factors are conceptualized as the opposite of a known risk factor (e.g., impulsivity vs. self-control; stress vs. coping). Other risk factors are the absence of a behavior or characteristic, and therefore, the protective factor is the presence of the behavior or characteristic (e.g., intimate relationships, participating in prosocial leisure or recreational activities). Some protective factors (or, rather, the lack thereof) have been found to predict future violence even when risk level was controlled.
Risk Assessment Measures
Static-Only Measures
Violence Risk Appraisal Guide–Revised (VRAG-R)
The VRAG-R is a 12-item actuarial measure of static factors of violence risk. Items include those related to psychosocial history (i.e., elementary school maladjustment, lived with both parents, and marital status), criminal history (i.e., admission to corrections, failure on conditional release, nonviolent criminal history, violent criminal history, sexual offending, and age at index offense), and psychopathology (i.e., antisociality, conduct disorder, and substance abuse). Each item is given a weighted value based on the item’s ability to discriminate between those who recidivate and those who do not recidivate. Total scores on the VRAG-R range from −25 to +27; based on this score, offenders are placed into one of nine bins with higher numbered bins associated with a stronger likelihood of violence.
Psychopathy Checklist–Revised (PCL-R)
The PCL-R is a 20-item measure of static factors examining an individual’s history of Interpersonal, Affective, Lifestyle, and Antisocial characteristics related to psychopathy. The PCL-R is measured on a 3-point scale (0–2) with a traditional cut score of 30, reflective of psychopathy. Although the PCL-R was not initially developed as a measure of violence risk, research has consistently linked the two concepts. In addition to the PCL-R, there is a shortened version and a version for adolescents, both of which are often used in violence risk assessments.
Static-99R
The Static-99R is a 10-item actuarial risk assessment measure of static risk factors related to sexual recidivism. Scores on the Static-99R range from −3 to 12, which places the individual in one of four risk categories: low (−3 to 1), moderate– low (2 to 3), moderate–high (4 to 5), and high (6 or higher). Research suggests that the Static-99 is one of the most commonly utilized measures of sexual recidivism risk among mental health practitioners, particularly those who conduct sexually violent predator evaluations for civil commitment proceedings. There is an updated version of this measure, the Static-2002R, which has additional items and separates individuals into one of five categories. Research is contradictory, however, on whether the 2002 version has any improved accuracy compared with its predecessor.
Rapid Risk Assessment for Sex Offense Recidivism
The Rapid Risk Assessment for Sex Offense Recidivism is a 4-item assessment tool for determining sexual recidivism risk among adult males who are known to have committed at minimum one sexual offense. The 4 items include the number of prior charges or convictions for sexual offenses, age upon release from prison or the anticipated opportunity to reoffend in the community, any unrelated victims (coded as yes or no), and any male victims (coded as yes or no). Scores on this measure range from 0 to 6.
Static and Dynamic Measures
Historical, Clinical, Risk Management-20 (HCR-20), Version 3
The HCR-20, Version 3 is a measure of violence risk that relies on structured professional judgment. It includes 20 items. Ten of these items (Historical) are static, and 10 (five Clinical and five Risk Management) are dynamic. The
Historical variables examine characteristics of the individual that are stable and unchanging (e.g., having a history of violence or antisocial behavior, history of problems with employment, and history of substance abuse). The Clinical variables include information about the individual’s current mental status (e.g., problems related to insight, symptoms of a major mental disorder, recent problems with violent ideation or instability) and the Risk Management variables examine environmental factors (e.g., problems related to their living situation, problems related to treatment or supervision). Each item is scored on a 3-point scale (0–2; absent, partially present, definitely present) with higher scores indicating the presence of the risk factor. A total score can be calculated (maximum score of 40), but the authors urge clinicians to integrate the variables into a final determination of violence risk (low, moderate, high).
Sexual Violence Risk-20 (SVR-20)
The Sexual Violence Risk-20 (SVR-20) is a 20-item checklist of both static and dynamic risk factors associated with future violence and SVR. Items on the Sexual Violence Risk-20 SVR-20 fall into one of three categories: Psychosocial Adjustment (i.e., sexual deviance, victim of child abuse, substance use problems, psychopathy, major mental illness, suicidal/homicidal ideation, and relationship problems), History of Sexual Offenses (i.e., high-density multiple types, physical harm, weapons/threats, escalation in frequency or severity, extreme minimization/denial, attitudes that support or condone), and Future Plans (i.e., lacks realistic plans and negative attitude toward intervention). Clinicians rate each item based on degree of risk for this defendant as low, moderate, or high.
Minnesota Sex Offender Screening Tool–Revised (MnSOST-R)
The Minnesota Sex Offender Screening Tool– Revised includes 16 items related to both static (e.g., age, presence of adolescent antisocial behavior, nonsexual and sexual criminal history) and dynamic risk factors (e.g., current/ recent employment instability, current/recent substance abuse, and current/recent sex offender treatment) for sexual recidivism. Scores on the MnSOST-R range from −14 to 30. The total score results in the assignment to one of three levels of risk: Level 1 (low risk, scores of 3 and lower), Level 2 (moderate risk, scores of 4 to 7), or Level 3 (high risk, scores of 8 or higher). The developers of the MnSOST-R also developed a subset of Level 3 (scores of 13 or higher), in which it is recommended individuals with these scores should be referred for commitment proceedings. The MnSOST-R is commonly used among mental health practitioners conducting sexually violent predator evaluations.
Dynamic Only Measures
Structured Assessment of Protective Factors
The Structured Assessment of Protective Factors for violence risk focuses on dynamic, protective factors affecting violence risk (only 2 items on the measure examine static factors). It was created as a complement to the HCR-20’s risk factors and is used together with that measure. Similar to the scoring design of the HCR20, the Structured Assessment of Protective Factors has 17 items that are scored on a 3-point scale (0–2) with higher scores indicating the presence of protective factor. These 17-items are categorized under one of three scales: Internal Factors, Motivational Factors, and External Factors. Clinicians can also score a Final Protection Judgment (low, moderate, high) based on an integration of the protective factors. Furthermore, clinicians can examine an integrated Final Risk Judgment by integrating the Final Protection Judgment with the HCR-20 risk scores to determine a risk of future violent behavior.
Use of Static and Dynamic Risk Factors
The aforementioned measures are the most empirically supported and do not constitute every violence or sexual recidivism risk measure that exists. Based on available research, the HCR-20, VRAG-R, and the Static-99 appear to have the most empirical support. Other measures, such as the Classification of Violence Risk and the Short-Term Assessment of Risk and Treatability, exist but have less research support or have been validated using a limited sample population.
Meta-analyses have found few differences in violence risk prediction across measures. Overall, the aforementioned measures moderately predict violence risk. While this suggests that each of these measures is interchangeable, it is important for clinicians to keep in mind the individual and the situation when choosing a risk assessment measure. Furthermore, although research has not found differences between measures of only static factors versus measures that include dynamic factors, these measures serve differing purposes. Measures that include dynamic factors allow clinicians to assess for changes in risk level over time or based on situation, whereas measures using only static factors do not allow for change. In conducting comprehensive risk assessments, clinicians are on solid ground to rely on both historical and dynamic variables. Such an approach provides the most valid predictions of risk.
References:
- Bonta J., & Andrews, D. A. (2016). The Psychology of Criminal Conduct. New York, NY: Routledge.
- Jaber, F. S., & Mahmoud, K. F. (2013). Risk tools for the prediction of violence: “VRAG, HCR-20, PCL-R.” Journal of Psychiatric and Mental Health Nursing, 22, 133–141. doi:10.1111/jpm.12102
- Miller, C. S., Kimonis, E. R., Otto, R. K., Kline, S. M., & Wasserman, A. L. (2012). Reliability of risk assessment measures used in sexually violent predator proceedings. Psychological Assessment, 24(4), 944–953. doi:10.1037/a0028411
- Skeem, J. L., & Monahan, J. (2011, March). Current directions in violence risk assessment. Public Law and Legal Theory Research Paper Series No. 2011– 13. Retrieved from http://ssrn.com/abstract=1793193
- Vitacco, M. J., Gonsalves, V., Tomony, J., Smith, B. E. R., & Lishner, D. A. (2012). Can standardized measures of risk predict inpatient violence? Combining static and dynamic variables to improve accuracy. Criminal Justice and Behavior, 39, 589–606. doi:10.1177/0093854812436786
- Vogel, V., Robbe, M. V., & Douglas, K. S. (2013). Risk factors and protective factors: A two-sided dynamic approach to violence risk assessment. The Journal of Forensic Psychiatry & Psychology, 24, 440–457. doi:10.1080/14789949.2013.818162
- Yang, M., Wong, S. C. P., & Coid, J. (2010). The efficacy of violence prediction: A meta-analytic comparison of nine risk assessment tool. Psychological Bulletin, 136(5), 740–767. doi:10.1037/a0020473