Gender Differences in Behavioral Outcomes among Children in the Family Connections Program

Background & Purpose: Poverty and exposure to community violence are noted to impact behavioral outcomes among urban youth, and recent studies suggest that urban African American boys may be of particular concern due to these concomitant neighborhood risks (Gorman-Smith, Henry & Tolan, 2004; Levanthal & Brooks-Gunn, 2003). The purpose of this study was to explore differences in internalizing and externalizing problem behaviors among urban, African American boys and girls served by the Family Connections (FC) program.

Method: Secondary data analyses explored behavioral outcomes among children from 154 families who met risk criteria for neglect, and who had at least one child experiencing behavioral problems. Families were randomly assigned to receive 3- or 9-month interventions. Using mixed model analyses of variance (ANOVA) with repeated measures, differences in child behavior outcomes between boys and girls were examined among the 128 families for whom pre and post-data were available.

Results: For internalizing behaviors, a significant Time x Treatment Group (F (1, 124) = 6.795, p < .05, Eta-squared = .05) interaction was obtained, though the effect for this interaction was weak. CBCL internalizing raw scores decreased for the FC 3-month group from Time 1 (M = 8.76) to Time 2 (M = 8.30) and for the FC 9-month group from Time 1 (M = 9.05) to Time 2 (M = 6.76). A trend Time x Gender (F (1, 124) = 3.797, p = .054, Eta-squared = .03) interaction was observed indicating that for the FC 3-month and FC 9-month intervention collectively, boys and girls' CBCL internalizing raw scores improved over time [(Boys: Time 1: M = 9.50; Time 2: M = 7.10), (Girls: Time 1: M = 8.31; Time 2: M = 7.96)].

For externalizing behaviors, significant Time x Treatment Group (F (1,124) = 9.82, p < .05, Eta-squared = .07) and Time x Gender (F (1, 124) = 6.20, p < .05, Eta-squared = .05) interactions were obtained, though the effects for each interaction were weak based on Eta-squared values. CBCL externalizing raw scores decreased for the FC 3-month group from Time 1 (M = 16.09) to Time 2 (M = 14.31) and for the FC 9-month group from Time 1 (M = 18.23) to Time 2 (M = 15.03). For the FC 3-month and FC 9-month intervention collectively, boys and girls' CBCL externalizing raw scores improved over time [(Boys: Time 1: M = 19.81; Time 2: M = 15.34), (Girls: Time 1: M = 14.51; Time 2: M = 14)]. Across intervention time points, boys appeared to experience larger decreases in internalizing and externalizing behaviors than girls.

Conclusions & Implications: Developing programs that effectively reduce child behavior problems is important because of the relationship between early onset of child behavioral problems on future delinquency and substance abuse. Findings highlight the promise of an early intervention targeting problem behaviors among urban, African American children, especially boys.