Children who witness violence, and parent report of children's behavior.

OBJECTIVES: To examine how much distress children report in response to violence that they have witnessed and how this is associated with parental reports of children's behavior.
METHODS: As part of a study of in utero exposure to cocaine, children completed the Levonn interview for assessing children's symptoms of distress in response to witnessing violence. The children's caregivers completed the Exposure to Violence Interview (EVI), a caretaker-report measure of the child's exposure to violent events during the last 12 months. The EVI was analyzed as a 3-level variable: no exposure, low exposure, and high exposure. The caregivers also completed the Children's Behavior Checklist (CBCL).
RESULTS: Of 94 six-year-old children, 58% had no exposure to violence, 36% had low exposure to violence, and 6% had high exposure to violence, according to caretaker reports. The children's median+/-SD Levonn score was 64 (SD +/- 19.3). The mean SD +/- CBCL total T-score was 53 (SD +/- 10.2). In multiple regression analyses with gender, low and high exposure on EVI, Levonn, and prenatal cocaine exposure status as predictors, the Levonn score explained 4.8% of total variance in children's CBCL internalizing scores, 9.1% of the total variance in CBCL externalizing score, and 12.2% of the total variance in CBCL total score (P =.04, P =.004, and P<.001, respectively).
CONCLUSIONS: After accounting for the caretaker's report of the level of the child's exposure to violence, the child's own report significantly increased the amount of variance in predicting child behavior problems with the CBCL. These findings indicate that clinicians and researchers should elicit children's own accounts of exposure to violence in addition to the caretakers' when attempting to understand children's behavior.