Several authors have tried to address the problem of impact, effect, and relationship amongst the collection of coupling measures by introducing frameworks to characterize different approaches to coupling and the relative strengths of these, although, on their own, none of the frameworks could be considered comprehensive. There are three existing and quite different frameworks for object-oriented coupling. First, Eder et al. identify three different types of relationships. These relationships, interaction relationships between methods, component relationships between classes, and inheritance between classes, are then used to derive different dimensions of coupling which are classified according to different strengths. Second, Hitz and Montazeri approach coupling by deriving two different types of coupling: object level coupling and class level coupling which are determined by the state of an object and the state of an objects implementation respectively. Again different strengths of coupling are proposed. And third, Briand et al. (1997) constitute coupling as interactions between classes. The strength of the coupling is determined by the type of the interaction, the relationship between the classes, and the interactions locus of impact. As none of the frameworks has been used to characterize existing measures to the different dimensions of coupling identified, the negative aspects highlighted above are still very common ones.
