When working within the criminal justice field it is important to be able to accurately assess the likelihood that an offender will cause problems while confined, reoffend once released, or fail to complete post-release requirements as well as a number of other situations that regularly arise with offenders. Static and dynamic risk assessment instruments can both be used to assess these likelihoods as well as many others. Although static and dynamic risk assessment instruments have a number of similarities and differences they both also have their strengths and weaknesses.
A static factor is one that does not change over time. For example, severity of offense, escape history, and prior felonies are all considered static factors. There have been a number of instruments found valid when predicting offender misconduct while institutionalized that rely on static factors (Van Voorhis, Braswell & Lester, 2007). However, in current years is has been shown that these instruments are not valid in regards to female prisoners and thus are only recommended to be used with male prisoners (Van Voorhis, Salisbury, Wright & Bauman, 2008). In order to compensate for the unchanging nature of static factors, reclassification systems use additional characteristics or deemphasize characteristics used during the initial classification in order to more accurately place offenders (Van Voorhis et al., 2007).
A dynamic factor, in contrast, is one that changes over time. These factors can, and should, be reassessed after a set period of time in order to more accurately place offenders where they will receive the correct levels of supervision or assistance. Dynamic factors include such items as educational level, drug abuse, and financial situations as an example. Instruments that rely on dynamic factors, at least in part, are able to reassess offenders without having to change the instrument from one assessment to the next. By utilizing these and other dynamic factors an offender can be reassessed more easily and efficiently. Additionally, these instruments have been found valid for both male and female offenders, although there are instruments available that have been designed specifically for female offenders (Van Voorhis et al., 2007).
In order to give practitioners a better understanding of risk assessment, Bonta (2002) has written 10 guidelines for the selection and use of risk instruments. Bonta does not distinguish between static and dynamic risk assessments and instead instructs in how to choose appropriate assessments based on the reasons for the assessments and the end goals of the assessment. The instruments that are then chosen based on these guidelines may include both static and dynamic factors or rely solely on one over the other.
The first guideline suggests that offender risk assessment should be based on actuarial measures of risk and not based solely on clinical measures of risk. Bonta (2002) describes several risk scales based on actuarial measures of risk that have been specifically designed to predict recidivism but rarely used by members of the criminal justice field; for example, the Psychopathy Checklist-Revised (PCL-R), the Level of Service Inventory-Revised (LSI-R), and the Violence Risk Appraisal Guide (VRAG). These instruments are comprised of both static and dynamic risk assessment measures and have been shown to accurately reflect recidivism risk even among special offender populations, such as the mentally ill or sexual offenders (Andrews & Bonta, 2007; Bonta, 2002).
Predictive validity should be demonstrated by the risk assessment used. This guideline seems quite straight forward, a risk assessment strategy should not be used unless is has been demonstrated that is does in fact assess what it was designed to assess. For example, the HCR-20 was designed to be used with offender populations but there have been very few studies completed that show the predictive validity of this instrument. Additionally, it is important that these instruments are shown to have predictive validity in regards to offender populations if they are expected to be effective. It is for this reason that the instruments used should be derived from relevant theory (Bonta, 2002).
In addition, it is important that the instruments used are relevant to criminal behavior. Although it would be possible to administer an instrument to a number of offenders and be able to predict future behavior from that instrument it is important that criminal justice employees remember that behaviors being measured should have a direct connection with criminal behavior if they are to be effective. Furthermore, it is also important that criminogenic need factors be assessed. Once these criminogenic need factors have been acknowledged and corrected, corrections in criminal behavior are more likely to occur (Bonta, 2002).
Multiple domains should also be examined when trying to determine the causes of criminal behavior. Since criminal behavior is caused from a wide variety of situations and circumstances, there should be a number of static and dynamic factors used when determining an appropriate treatment course or the proper way in which to detain an offender. In order to match treatment and detainment with an offenders unique cognitive, personality, and sociocultural characteristics it is important to limit the tests conducted to those that assess responsivity. Examples of such tests include the I-Level and the CM (Bonta, 2002).
Just as different domains should be examined, these domains should be tested in a number of ways in order to limit the error that is associated with any one instrument. Within the criminal justice field there are four popular testing formats, each with a number of tests associated with them that are popular within the field. A popular paper and pencil test is the I-Level test used for general classification purposes. Interview-based tests include the LSI-R and the HCR-20. The Adult Internal Management System is one example of a behavioral measure while the VRAG and STATIC-99 utilize the file extraction method. Using just one instrument or several instruments utilizing the same testing method introduce error into the testing that could alter the results of the testing (Bonta, 2002).
The last two guidelines are focused more on reminding practitioners to use tests responsibly by exercising professional responsibility and by conducting oneself in a respectful manner. Bonta feels that it is important that criminal justice workers should remember the importance of the tests being conducted and that they not only affect the offender but also the safety of the staff and officers where the offender is placed. Additionally, Bonta points out that the instruments used should adhere to the principle of the least restrictive alternative and should not interfere or put undue strain on the offender if it is not absolutely necessary (Bonta, 2002).
Regardless of whether the risk assessment instrument being used is designed mainly using static or dynamic risk factors it is important that the instrument being used was designed to predict the behavior in question and that the validity of the instrument has been shown for the population it is to be used on. Although static and dynamic risk factors differ from each other in what they can predict and how they are to be used, both have been shown to work well with certain groups when used properly.
Andrews, D. & Bonta, J. (2003). The Psychology of Criminal Conduct, Cincinnati, OH: Anderson Press, Ch 6.
Bonta, J. (2002). Offender Risk Assessment: Guidelines for Selection and Use. Criminal Justice and Behavior, 29: 355-379.
Van Voorhis, P., Braswell, M., & Lester, D. (2007). Correctional Counseling and Rehabilitation, 6th ed., Cincinnati, OH: Anderson Press., Ch 7.
Van Voorhis, P., Salisbury, E., Wright, E., & Bauman, A. (2008). Achieving Accurate Pictures of Risk and Identifying Gender-Responsive Needs: Two New Assessments for Women Offenders. Washinton, D.C.: United States Department of Justice: National Institute of Corrections.