Analysis of Family Research Designs

Publisher Website
Google Scholar
Abstract
This article presents both a conceptual model and an analytical method that can be used in the evaluation of round-robin family research data. The general model proposed is a variant of the social relations model (Kenny & La Voie, 1984). This model is used to partition family data into individual, dyadic, and family effects and permits examination of several types of interdependence between family members. At the individual level, actor and partner effects for each family role (mother, for example) can be estimated. At the dyadic level, relationship effects can be estimated; and at the family level, mean differences between families can be estimated. By correlating these different effects across family roles, this technique allows researchers to examine several types of interdependence within families. The analytical method presented is confirmatory factor analysis. Strengths and weaknesses of this model and analytical strategy are discussed.