Purpose: This study utilized two primary measurement instruments aimed at assessing family functioning and parenting practices and evaluated them for their usefulness as a measure of family attachment. There was also interest in the separateness of the measures of family functioning and parenting practices that remained separate upon analysis despite their viability as a possible measure of attachment. The refined attachment measure was then tested for its predictability in traits and behaviors in children. The study is based on the foundation of the Social Cognitive Learning Theory and Ecological Theory combined as an explanation of attachment qualities that impact behaviors in children.

Method: Data were gathered from a NIAAA funded (RO1 AA1164701A1) bi-national (United States and Canada) family-based prevention program (N = 674). Families with at least one child between the ages of 9 and 12 and a parent with an alcohol problem were recruited into the study. Family functioning, measured by parent reports on the Family Assessment Measure (FAM III), and parenting practices, measured by the Alabama Parenting Scales, were assessed and reconfigured as a measure of attachment. Exploratory factor analyses were used to identify the potential dimensions of family attachment. Reliabilities and multiple regressions were performed to assess the refined measurement instrument. Dependent variables consisted of child reports of problem behaviors, depression, victimization, coping strategies (including attention-seeking and problem-solving), and social skills. Separate forward stepping ordinary least squares regressions were run for each of the child behaviors.

Results: The final factor analysis identified six factors. Cronbach's alpha coefficients for these six dimensions were satisfactory: relationship = .924; involvement = .853; temperament = 758; indifference = .765; discipline = .688; and happiness = .735. Correlations between factors were small (.18) to moderate (.53). Results from the regression analyses indicated the following: indifference was positively associated with conduct problems, oppositional behaviors, and victimization, and negatively associated with problem solving coping and social skills. Temperament was negatively associated with depression. Involvement was negatively associated with oppositional behavior as was happiness. Relationship was positively associated with overall coping skills. There were no significant relationships for attention seeking coping or self esteem.

Implications: The factor analysis resulted in six relatively independent dimensions of family attachment. It is also of interest that the family functioning and parenting items remained independent of one another, falling out on separate factors. Regression results suggest that for this sample of children of alcoholics, the reports of problem behaviors by children were not as strongly related to what might be considered family attachment as might have been expected. However, parental indifference, reflecting characteristics of parental monitoring, showed a steady relationship with the child behaviors. This is consistent with prior work on predictors of problem behaviors in children. Furthermore, attachment is a difficult construct to measure and the results suggest that clinicians who are looking to identify at-risk children need to consider other factors in addition to family attachment. Practitioners and researchers may be missing other factors that are important in predicting child behaviors if they solely assess for attachment.