Why don't family measures of cohesion and control behave the way they're supposed to?

This paper examines the difficulties in family assessment research with the two central constructs of cohesion and control. First the status of these constructs in the literature is reviewed, including evidence for their convergent and discriminant validity. Then problems stemming from unarticulated theory, inadequate definitions, and lack of fit between definitions and instrumentation are described. Following this critique, a conceptual reappraisal of the constructs of cohesion and control is offered, based on the family FIRO model, and a framework for making sense of the problems encountered thus far in assessing these domains of family interaction is suggested.