Guidelines for Conducting Positivist Case Study Research in Information Systems

Abstract

The case study research approach is widely used in a number of different ways within the information systems community. This paper focuses on positivist, deductive case study research in information systems. It provides clear definitions of important concepts in positivist case study research and illustrates these with an example research study. A critical analysis of the conduct and outcomes of two recently published positivist case studies is reported. One is a multiple case study that validated concepts in a framework for viewpoint development in requirements definition. The other is a single case study that examined the role of social enablers in enterprise resource planning systems implementation. A number of guidelines for successfully undertaking positivist case study research are identified including developing a clear understanding of key concepts and assumptions within the positivist paradigm; providing clear and unambiguous definitions of the units and interactions when using any theory; carefully defining the boundary of the theory used in the case study; using hypotheses rather than propositions in the empirical testing of theory; using fuzzy or probabilistic propositions in recognising that reality can never be perfectly known; selecting case studies carefully, particularly single case studies; and recognising that generalisation from positivist, single case studies is inherently different from generalisation from single experiments. When properly undertaken, positivist, deductive case study research is a valuable research approach for information systems researchers, particularly when used within pluralist research programs that use a number of different research approaches from different paradigms.