Substantiation as a multitier process: the results of a NIS-3 analysis.

Abstract

Previous studies on child maltreatment reporting have focused mainly on one level of substantiation. This article analyzes factors influencing the multitiered substantiation process. The 1993 Third National Incidence Study (NIS-3) data of substantiated and non-substantiated reported incidents (N=7,263) of maltreatment were analyzed. Substantiation was classified into three categories: unfounded, indicated, and founded. Independent variables included demographic characteristics, case-processing variables, and maltreatment characteristics. Bivariate and multiple logistic regression (MLR) analyses were calculated to determine whether demographic and case processing variables predicted unfounded or founded/indicated dispositions. Second-level analysis examined demographic, case processing, and maltreatment characteristics as predictors of founded or indicated status. These results showed that 60.2% of CPS investigations conducted were evaluated as unfounded, about 22% were categorized as founded, and 17% were classified as indicated. In the MLR analysis for the first level of substantiation, case processing variables were highly significant predictors of founded/indicated status. In the second-level substantiation MLR model, cases in the mid-range income level (dollars 15,000-29,999) had a lower probability (adjusted OR = .58, p = .02) of being founded than those of less than dollars 15,000, and reports involving Hispanic children (OR = 3.04, p = .05) were more likely than the "all other" race-ethnic social classification to have been substantiated as founded. This analysis of NIS-3 data suggests that a three-tiered rather than a two-tiered system is a more accurate representation of the CPS substantiation process. Further analysis of substantiation patterns is required to provide a basis for developing more effective investigation systems.