Integrating models of interdependence with treatment evaluations in marital therapy research.

Clinical intervention may not be the only source of change in the behavior of husbands and wives during marital therapy. Spouses may continue to influence each other, though perhaps in ways different than they did before treatment. Viewed as a short longitudinal study of marital interaction, a pretest-postte st trial of marital therapy can provide information on the short-term processes of interpersonal influence or interdependen ce that occur over the course of treatment. Moreover, interpersonal processes and treatment processes may combine to produce unique clinical outcomes. This article presents 3 structural equation models of marital interdependence that can be applied to pretest-posttest data from clinical trials of marital therapy. Modifications of these models that include a treatment variable are also presented. These models provide tests for treatment effects that are unique to interventions into relationship systems. At least since the writings of Sullivan (1953), the psychological well-being of individuals has been considered within the context of close interpersonal relationships. The key insight of the early thinkers was that the behavior of closely related individuals can be interdependent, such that a change in one member of such a pair (or system) entails an accommodating change in one or more partners (Bowen, 1960; Bowlby, 1949; Jackson, 1957). Moreover, resistance to change within relationship systems can prevent individual development. Such considerations gave rise to a burgeoning array of marital and family therapies. Whereas some therapists have considered the complexity of interpersonal systems to be unamenable to scientific inquiry, others have attempted to demonstrate that