Concurrent child history and contextual predictors of children's internalizing and externalizing behavior problems in foster care.

This study contributes to current research on the behavior problems of children in foster care by analyzing a more comprehensive set of concurrent child history and contextual predictors. Kinship home status and sibling status (i.e., whether the sibling was a biological sibling to the foster child) were evaluated as moderators of significant associations. Data were collected at the baseline of a foster parent training intervention program prior to any intervention services using parent phone interviews ( N = 310, 51.6% male, M age = 7.57 years). Two linear hierarchical regressions were used to evaluate each set of predictors' association with behavior problems as well as each individual predictor's contribution. Results indicated that as a set, the contextual variables predicted a significant and unique amount of variability in the child's internalizing and externalizing behavior scores, but the child history variables did not. Specifically, the child's placement in a non-kinship home, being in a non-ethnically matched child-parent pair, higher parent stress scores, a greater number of prior group home placements, and higher internalizing behavior scores for the child predicted higher child externalizing scores. Higher parent stress scores, higher focal sibling externalizing behavior scores, and higher externalizing behavior scores for the child predicted higher internalizing scores for the child. The association between focal sibling externalizing behavior scores and child internalizing scores was moderated by kinship home status, such that there was a stronger association between the focal sibling's externalizing score and the child's internalizing score if the child was in a kinship compared to a non-kinship home. Implications for intervention services are discussed, particularly the importance of assessing the child's foster home environment when addressing the child's behavior problems.