Neighborhood conditions, father involvement, parenting competence, and behavior problems in a sample of children at risk for neglect: A structural equation model

Abstract

Internalizing and externalizing behaviors in children have been associated with continued patterns of anti-social behavior into adolescence and adulthood, including later risk for arrest (Patterson et al., 1992) and adult substance abuse (Ferdinand, Blum, & Verhulst, 2001). The primary aim of this dissertation was to broaden the understanding of neighborhood and family factors associated with child behavior problems in a sample of children residing in impoverished urban neighborhoods who were at risk for child neglect. This study used a cross sectional design to conduct secondary data analyses on 199 families who were served by the Family Connections program. The primary aim was to test a structural model explicating the relationships among neighborhood characteristics, perceived social capital, father involvement, parenting competence, and child behavior problems. Structural equation modeling (SEM) with robust maximum likelihood (MLM) estimation was used to test two structural regression models. Good model fit was obtained for both the internalizing behavior (Santorra-Bentler Scaled χ2= 57.9753 (df=30); χ2//df=1.9; RMSEA= .06; SRMR=.07; GFI=.94; NNFI=.91; CFI=.94) and externalizing behavior (Santorra-Bentler Scaled χ2= 58.8362 (df=30); χ2/df= 1.96; RMSEA= .06; SRMR=.07; GFI=.93; NNFI=.90; CFI=.93) models. Examination of path estimates revealed an indirect effect of neighborhood and social capital constructs on child behavior via parenting, but no direct effect of neighborhood or social capital on either internalizing or externalizing child behavior. Father involvement was not associated with parenting sense of competence but was related to externalizing behavior at a level that was statistically significant yet small in magnitude, and unrelated to child internalizing behavior in this analysis. The relationship between parenting competence and child behavior problems was significant for both models. This relationship was the largest in magnitude in the analysis. The combined effects of neighborhood conditions, social capital, father involvement, and parent competence for accounted for 24% (for internalizing) and 20% (for externalizing) of the variance with most of this observed effect deriving from the parenting construct.