Emerging methods
pp. 715-727
in: Pauline Boss, Ralph LaRossa, Walter R. Schumm, Suzanne Steinmetz (eds), Sourcebook of family theories and methods, Berlin, Springer, 1993Abstract
Theory and methods in the social sciences, particularly statistical methods, are inextricably intertwined. On one hand, theoretical frameworks, as isomorphic and homomorphic images of reality, determine the research questions that can be asked and the nature of data gathered and require particular statistical procedures for description and testing. On the other hand, statistical procedures, by specifying a specific mathematical relationship between variables or concepts, influence the nature and form of observed empirical regularities used to develop theoretical propositions, limit the types of data that can be analyzed, and restrict the sorts of theoretical propositions that can be tested. From an extreme position some argue that statistical methods tell us little about "reality" since the conclusions are built in to the original definitions, assumptions, and rules the researcher employed.