Behavior modification, although the most articulately averse to formal statistics, is the main area that has given concerted attention to single subject methodology. Several books have been written, all of which can be read with profit, as they attack real problems that are also important in other areas. These books, like the present book, represent an empirical direction in design and analysis. Methodology consists in cumulative experimental procedure, not static precepts, but evolving knowledge systems. e
Each area faces its own special problems and needs to develop its own single subject methodology. Methodology is a continual concern of active investigators, and beginners in a field can profit from attention to issues of design and procedure in published articles. Much methodology, however, remains task- and area-specific lore. Efforts to crystallize such lore and make it useful to workers in other areas would be helpful.
Methodology is a bad word to many. Most investigators are truly concerned with methods, of course, but the term methodology suggests a dogmatic stance on standardization of procedure and correct data analysis. It connotes involvement in niceties and complexities of apparatus and especially statistics that are generally barren, often useless digressions, sometimes active hindrances to productive inquiry.
Properly considered, however, methodology is an organic part of substantive inquiry. Necessarily so, for the validity of methods derives from the empirical results that they bring in.… Knowledge is not divorced from the methods by which it was acquired; those methods themselves constitute an integral part of knowledge. (Anderson, 1982, p. 349.)
For helpful comments on drafts of this chapter, I am indebted to Ted Carr, Joe Farley, Etienne Mullet, Laura Scbreibman, Saul Sternberg, Ben Williams, and Wendy Williams.