Innovative Techniques: Siren or Rose

Abstract

ion. Secondly, it forces us to specify exactly what different kinds of theoretical relationships are used. Thirdly, such a classification helps us recognize the need to clarify situations in which our predictions will hold. Finally, and perhaps most important, such a classification scheme forces us to identify clearly the frame of reference for the theoretical domain we have constructed. Stinchcombe (1973) laid out a preliminary sketch of this form of classification scheme. He included four categories: a set of objects, a set of operations, a set of assumptions, and a set of environments. Set of objects. The first type of elements in a scientific theory are its conceptual objects or terms. These are divided into two groups. The first group contains the abstract level terminology of concepts, constructs, variables, and operational definitions. Over the last few years in family studies, the emphasis has been to clarify concepts and develop variables (Holman and Burr, 1980). The second group contains the concrete level terminology of real objects, their ascribed attributes, and our measuring instrumentation. A complete scientific theory contains objects at both the abstract and concrete levels. Set ofoperations. In mathematics an operation is the manner or way two or more objects or terms are related or connected to each November 1982 JOURNAL OF MARRIAGE AND THE FAMILY 877 This content downloaded from 157.55.39.35 on Thu, 01 Sep 2016 05:05:57 UTC All use subject to http://about.jstor.org/terms other. Operations are divided into three groups. The first group contains the horizontal relationships that operate between objects at the same level of abstraction. Examples are: (a) a positive relationship between life satisfaction and income, (b) the correlation between two measurements or (c) a legal marriage between a man and a woman. The second group contains the definitional operations that interlink terms either horizontally or vertically. Examples are: (a) marital adaptation is a type of marital quality, (b) the Rotter Index is a measure for locus of control or (c) quality is defined as. . . . The third group contains the logical operations that vertically link the horizontal class of relationships. Examples are: (a) abstracting a process from observing the behavior of real objects, (b) a form of deductive inference such as Aristotelian logic or (c) the fit between the data analysis strategy and the hypothesis to be tested. A complete scientific theory has all three types of operations. Set of assumptions. One of the most difficult tasks in theory construction has been the articulation of our assumptions. As the histories of our conceptual frameworks show, it is exceedingly frustrating to try mapping out the set of givens in such specific detail that either the whole set of deductions is well-grounded or the set of inductions can be stated with completeness. Thus, with rare exceptions like Volume II of Burr et al.'s (1979) Contemporary Theories About the Family, assumptions underlying our models or hypotheses to be tested are not stated. However, a complete and valid theory does contain assumptions even if unstated, and the need to specify those assumptions for evaluating the theory can never be stressed strongly enough (Blalock, 1979; Ruano et al., 1969; Shnell, 1975). Set of environments. In mathematics, philosophy and religion there are many theories that are internally consistent, draw self-evident conclusions, and are constructed exquisitely; but they are not scientific. A theory in each area includes all the elements discussed above under objects, operations and assumptions. What they lack is the specification of the real situations or environments in which the hypothesized relationships will operate as theorized. Without identifying such conditions, our theories are open to claims of uselessness or mere speculation. Thus in the construction of our theories it is important to specify conditions such as which types of families, in which situations, at what point in history, for what period of time, given what special constraints or initial conditions. A good example of specifying environments is found in Bahr (1979:618and 629), where he identified the conditions under which different models of delinquency might operate. One wonders, for instance, if the debate over the cause of schizophrenia among proponents of family, genetic, environmental, psychological or biological explanations possibly could be clarified if each researcher more clearly specified the environments and conditions under which his or her theory might explain some cases of schizophrenia rather than all