ABSTRACT
When studying family relationships, researchers often are interested in relational phenomena at an individual level, such as whether an individual child thrives in a particular family context; relational phenomena at the dyadic level, such as whether the mother is more committed than the father to the relationship; and relational phenomena at the familial level, such as whether some families are more satisfied with their relationships than other families. These three levels of relational phenomena, that is, individuals, dyadic relationships, and groups have to be distinguished and are reciprocally related within the family system (Hinde, 1997; Laursen & Bukowski, 1997). However, their theoretical and methodological interdependence represents a challenge for empirical investigation. The family as a group includes the individual family members and relationships, but these parts combine in such a way that new structures emerge which have their own meaning and are more than a collection of its members and relationships and which may affect relationships and the relationship partners (Reis, Collins, & Berscheid, 2000). Both statistical methods and theoretical frameworks are needed that allow us to disentangle the three levels and the influence they exert on each other and to capture the truly social nature of family relationships.
