Examination of the Predictive Validity of Risk Assessment Screening

Child Protective Services ( CPS) have been under growing pressure to implement risk assessment systems over the past ten years (Dueck, H. J., English, D.J., Depanfilis, and Moote, G.T. (1993). This emphasis has come largely from the increasing number of reported allegations of child abuse and neglect (Doueck, et. al.) However, increasingly individuals involved in custody battles use the CPS reporting system to retaliate against one another. Because these reported cases require the action of CPS, the increase in allegations has overburdened the system. As a result, resources have diminished and challenges to screeners to make accurate risk assessments have been stretched (Doueck, et.al.). The resultant strain on the protective services system has Jead some authors to question whether the system is capable of helping those children who are in most need (Wexler, R., 1990, Doueck, H.J., English, D.J. Depanfilis, D. and Moote, G.T. 1993). Consequently, a method of structuring the risk assessment process is needed to enable agencies to improve workload pressures through a comprehensive method of classifying cases by risk assessment to accurately predict future mistreatment raises serious questions and needs to be. further examined. The research question posed by this study is to what degree is risk assessment screening a prediction at the time of intake? The statement derived from this question is stated in a positive directional format:, Risk assessment screening will accurately predict the risk assessment findings verified during the intake process. A systematic random sample of 60 was made of county case records opened between July 1, 1993 and June 30, 1994 were reviewed at county Department of Social Services offices in San Bernardino and Rancho Cucamonga, California. The analysis of the study\u27s hypothesis and classification of additional findings made use of Kendall\u27s Tau-b (a method of cross tabulation analysis), Discriminate Analysis, Chi Square and Multiple Regression Analysis. The antecedent variable is risk assessment and the consequent variable is risk assessment intake code. The amount of association found between the consequent and antecedent variables was significant at the .001 (