Theoretical Inference and Generalisation Within the Case Study

Previous attempts to develop theory from qualitative case study research have employed the underlying principles of John Stuart Mill's system of logic, albeit in different ways (Mill, 1973). In his analysis of inductive proof, Mill proposed a set of formulae, or methods, to guide the process of induction. The two most important methods are considered to be the methods of agreement and difference. In the method of agreement, a constant association is sought across cases between the phenomena of interest, effectively identifying necessary conditions, but excluding extraneous conditions unconnected with the theory. Using the method of difference, however, the search is for sufficient conditions, those that must be present for a particular outcome to occur. Used together, the cannons comprise a stronger test than when employed individually, since the method of difference offers some security against the threat of alternative interpretations. Following the logic of both methods implies a search for conditions, which are always present when a particular phenomenon is present and which are never present when that phenomenon is absent. The basis for theoretical inference therefore rests on the comparison of similar and dissimilar cases as stated in Mill's methods of agreement and difference, and which finds its ultimate expression in experimental method (Cook & Campbell, 1979).