DISPOSITION OF CHILD PHYSICAL ABUSE REPORTS - REVIEW OF THE LITERATURE AND TEST OF A PREDICTIVE MODEL

Child protection programs have responded to the problem of insufficient staff to manage the influx of maltreatment reports through screening and triage mechanisms. Despite widespread implementation of such procedures, few studies have focused on developing empirical models that might help case-workers make these decisions. While the models that have been developed to date are superior to caseworker judgment for identifying report substantiation status, their predictive efficiency is not sufficient to warrant use. In this study, ordinal probit analysis was used to develop models to predict the disposition of physical abuse reports for 1447 families made to the public child protection program of a large eastern city using independent variables from four domains and an ordinal level dependent variable that indexed the intake caseworker's confidence that physical abuse had occurred. Analyses revealed that the best performing model involved the dichotomized dependent variable of ruled out and uncertain vs. indicated and confirmed.