Involuntary civil commitment: communicating with the court regarding "danger to other"

Beginning in the 1990s, researchers in the field of violence risk assessment recognized that even the most accurate and valid assessment could not assist fact-finders if the information was communicated in an unclear, imprecise, and/or incomplete manner (Monahan & Steadman, 1996; Schopp, 1996). The purpose of the current study was to investigate judges' opinions regarding the probative value of risk communication messages in civil commitment proceedings. Three types of risk communication messages were investigated: (1) description; (2) prediction; and (3) management. Additionally, the risk prediction model was investigated more in-depth by varying the format used to communicate the prediction: (1) categorical; (2) frequency; and (3) probabilistic. A national sample of 403 judges completed the study and via random selection, each received one of 10 risk vignettes. The vignettes described a situation in which a person underwent an evaluation for civil commitment based on the danger to other criteria; danger to self criteria was not at issue in this case. Risk messages and risk level were systematically varied. This study found that the risk models (i.e., description, prediction, and management) were viewed as equally probative. However, within the risk prediction model, categorical messages were viewed as the having the highest probative value. Results also indicated that of the three models, the risk prediction model led to rulings that are more strict than the other risk models, but no one risk prediction format resulted in higher restrictiveness. Results of this study can be used to provide mental health professionals some guidance as to how to present their data such that judges find the information easy to interpret and probative to their decisions.