Juvenile court personnel frequently must make judgments about juveniles' dangerousness in cases involving court referrals for alleged delinquencies. The statutes of more than two-thirds of the states include "danger to community" or "danger to others" as a legal criterion that must be met in order to allow pretrial detention of juveniles beyond a brief two or three day emergency period. Juvenile courts' dangerousness judgments are highly discretionary considering the importance of the judgments to the future welfare of juveniles as well as the interests and protection of communities. Not surprisingly, there has been considerable research interest in juvenile courts' judgments. However, past studies have told us little about courts' dangerousness judgments specifically. Most studies have concluded that accurate prediction of dangerousness is not possible. Dr. Grisso feels that the impasse that has been felt in the prediction of dangerousness is the result of too many methodological and conceptual problems that have manifested themselves in studies in this area. He hypothesizes that the failure of these attempts to yield workable predictors of violence could rest partly on an assumption that factors related to violence bear a linear relationship to violence and partly on the fact that studies have been based on inappropriate samples. In this study, he will investigate interactions among variables, using a cluster approach, and examine decisions made by a variety of actors in the juvenile justice system who provided data about their own decisions regarding dangerousness. Using data from a previous study which collected 432 cases (reported by a national sample of the same number of juvenile court personnel), he will develop prototypes (typical profiles) of juvenile cases perceived as "dangerous" by juvenile court personnel. Prototypes will be conceptualized as potential social schemata used by juvenile courts when processing case information to arrive at dangerousness judgments. This research will take a large and empirically derived domain of predictor variables and combine them in ways that reflect the practice of actual experienced decisionmakers. If juvenile case prototypes are found, it could greatly improve our capacity in future research to understand more clearly the nature of the dangerousness judgment process and to examine the validity of these judgments. Further, it would support the value of a social schemata approach for describing decisionmaking in this legal context, and would support generally the heuristic value of this theoretical approach in other decision research.
