| This study was undertaken to gain a more complete understanding of how the demographic and functional characteristics of children with or at risk of emotional and behavioral disorders are related to special education placement and mental health services use (N = 256). Parametric and nonparametric analytic techniques, including the use of the Lorenz curve, were used to determine an empirical method of distinguishing between ‘heavy’ and non-heavy users of behavioral health services. Classification and regression trees (CART) (Breiman et al., 1984) were used to develop models that describe the pathways, or interactions, of using the correlates of special education placement and mental health service use described in the literature. Main effects and interactive logistic regression techniques were used to validate the groups and the overall structure identified in the CART models. The investigation concluded with the analysis of patterns of service use through rank order analysis and chi-square tests of significance.; Findings suggest that in the prediction of heavy use of behavioral health services and supports, caregiver need in providing for their children, absences and home behavioral problems and symptoms were the most important predictors of high use of services and supports. Further analysis results suggest that ‘heavy users’ differ not only in the quantity of services used, but in the mix of services and supports. For special education placement, gender was the most important determinant of placement accompanied by reading scores, home problem behaviors and age of the child. Finally, the set of characteristics that places males at risk of special education placement differs from predictors of special education placement for girls. |
