It is commonly accepted in the computer industry that information systems projects are more likely to fail than not. Academic studies by e-government researchers provide the analytical findings that confirm this practitioner insight. Failure and success are subjective assessments that vary over time and with the standpoint of those making the judgment. Evaluation results are often contested, with the dispute based on political, legal or contractual matters – and even differing academic points of view.
This paper reviews the three main categories of diagnostic approach being used in the study of failed e-government projects – factoral analyses, systems approaches and interpretive studies. It shows that each category derives from a separate academic discipline, is based on differing theoretical constructs and entails a particular epistemological stance and research methodology. The story is told of the author's own experience in deciding on an appropriate research strategy for the study of a failed e-government project in Sri Lanka. Practical conclusions and recommendations are drawn to guide future research.