What is the most accurate violence risk assessment tool on the market today? This is a question asked every day by mental health, correctional, and legal professionals working across the globe. This is particularly the case in the United States, where surveys have estimated that over 80% of forensic psychologists use a structured instrument when conducting risk assessments. But you may be surprised to learn that there are over 400 risk assessment tools currently being used on six continents – all claiming to produce the highest rates of validity and reliability. Recent large-scale research has concluded that there is not a single risk assessment tool that consistently predicts future incidents of violence better than all others. Indeed, the risk assessment tool that is going to be most accurate for you is not the one with the best marketing campaign or even the one with the most studies published related to it. Rather it is the instrument with the strongest goodness-of-fit between how the tool was designed and how you use it. But how do you determine this goodness-of-fit? When deciding upon which violence risk assessment tool to adopt in practice, there are three key factors to take into consideration:

1. Population: Compare your average patient to the sample on which a risk assessment tool was normed, taking into consideration age (child, adolescent, adult), sex, race/ethnicity, nationality, offense history, and diagnostic group. For example, if an instrument was developed in a rural area of Canada on a predominantly Caucasian sample of men with an unclear diagnostic background, that risk assessment tool will likely not perform to its maximum ability in a unit serving predominantly minority female patients in downtown Chicago.
2. Setting: Compare the setting in which you are evaluating the average patient with the setting in which the normative sample was assessed. For example, if an instrument was developed using a group of patients evaluated upon admission to a forensic psychiatric facility, that risk assessment tool will likely not perform to its maximum ability when used by a parole board to make release decisions.
3. Outcome: Compare the outcome for which a risk assessment tool was designed with the outcome you are interested in predicting. For example, if an instrument was developed to evaluate the risk of general recidivism, that risk assessment tool will likely not perform to its maximum ability when used to predict sexual recidivism, specifically. Make sure to pay particularly close attention to the operational definition of the outcome in the risk assessment tool’s manual – instruments differ in terms of whether new arrests, charges, convictions, incarcerations, and/or self-reports of offending are included. Further, some risk assessment tools were developed for the prediction of intra-institutional infractions, whereas others were developed for the prediction of misconduct in the community.

Jerrod Brown, MA, MS, MS, MS, is the Treatment Director for Pathways Counseling Center, Inc. Pathways provides programs and services benefitting individuals impacted by mental illness and addictions. Jerrod is also the founder and CEO of the American Institute for the Advancement of Forensic Studies (AIAFS) and the lead developer and program director of an online graduate degree program in Forensic Mental Health from Concordia University, St. Paul, Minnesota. Jerrod is also currently pursuing his doctorate degree in psychology.

Jay P. Singh, Ph.D., is the founder of the Global Institute of Forensic Research (www.gifrinc.com), specializing in providing cutting-edge research, training, and software solutions to mental health, correctional, and legal professionals working in both general care and forensic settings around the world.