Out-of-home placement of children with severe handicaps: a comparison of approaches.

The relation of risk factors to out-of-home placement of children with severe and profound mental retardation was examined using both cross-sectional (n = 5,992) and 2-panel (n = 141) designs. Predictor variables included age, sex, ethnicity, adaptive behavior, maladaptive behavior, retardation level, blindness, deafness, and epilepsy. Multivariate analysis of variance indicated significant differences in the relative importance of predictor variables between sampling approaches. However, accuracy in predicting the children placed increased little in the large cross-sectional sample (model accuracy = 72.2%) over the smaller semi-longitudinal one (model accuracy = 70.2%). It appears that factors such as age, adaptive behavior, and maladaptive behavior may be overstated in large, cross-sectional studies of placement, and ethnicity may play a larger role in placement decisions than previous cross-sectional comparisons suggest.