Best predictor of problem behaviors in children: Is it the parent or the child?

Young children have significant rates of exposure to violence. In urban settings, the rate is as high as 80% (Mrug, et al., 2008). During pediatric visits, parents are frequently relied upon to report their child's social risk exposure and behavior problems. This study examined the sensitivity and specificity of child's self-reports and parent's reports of child's violence exposure to determine the best predictor of behavior problems. Participants included 409, primarily African American (86%) children, mean age 9.6 years (+ 1.1). The Children's Report of Exposure to Violence (CREV; Cooley, Turner and Beidel, 1995), Multicultural Events Schedule for Adolescents (MESA; Gonzales, et al., 1995), Youth Self Report (YSR; Achenbach & Dumenci, 2001), and Child Behavior Checklist (CBCL; Achenbach, 1991a) were used to identify violence and interpersonal stress exposure and behavior problems. A dichotomous behavior problem variable (0 = parent and child reported no clinically significant behaviors in child, 1 = parent and/or child reported significant behaviors) was created. For all violence exposure settings studied (community, family, peer), the child's reports of violence exposure were more sensitive in identifying clinically significant internalizing, externalizing, and total behaviors compared to the parent's reports (67-96%). For all three settings, the parent's reports of child's exposure to violence resulted in greater specificity in internalizing, externalizing, and total behaviors than the child's reports (61-91%). Given the child's report of violence exposure was more sensitive in identifying behavior problems, eliciting young children's reports of violence exposure as well as the parents' reports of child violence exposure is optimal.