You are not currently logged in.
Access your personal account or get JSTOR access through your library or other institution:
Prediction accuracy of the Washington and Illinois risk assessment instruments: An application of receiver operating characteristic curve analysis
Michael J. Camasso and Radha Jagannathan
Social Work Research
Vol. 19, No. 3 (SEPTEMBER 1995), pp. 174-183
Published by: Oxford University Press
Stable URL: http://www.jstor.org/stable/42657138
Page Count: 10
Preview not available
In recent years many child protective services (CPS) agencies have used structured risk assessment to streamline and rationalize case identification and management in a resource-depleted environment. Empirical evidence of risk assessment's capacity to predict child maltreatment has been limited to tests of single instruments. This article compares the predictive performances of the Illinois CANTS 17B and the Washington State Risk Matrix on a sample of New Jersey CPS cases using logistic regression and receiver operating characteristic (ROC) curve analysis. Both instruments predict case recidivism, closings, and substantiation with probabilities greater than chance. Moreover, in the prediction of case recidivism, the Washington instrument showed superiority at a statistically significant level. The implications of ROC analysis for child welfare decision makers and the relationship among test accuracy, problem prevalence, and diagnostic cutoff points are discussed.
Social Work Research © 1995 Oxford University Press