Web-Based Information Systems Development: An Empirical Study of Requirements Engineering Approaches

Abstract

No system is perfect in so far as requirements are always the subject of ongoing negotiation, and there will generally be a subset of the requirements that get pushed aside. This is especially true of the volatile environments that characterise Web-based information systems (WIS) development. A pragmatic resolution is to apply a requirements prioritisation methodology that aims to produce a best possible wish list under constrained circumstances. Herein are reported the findings of a case study of a WIS development project which uses such an approach. In practice, requirements engineering methodologies are not executed in the structured, methodical way advocated by researchers. This is not surprising, as the underlying philosophy of most of these methodologies is that systems development is a rational process, whereas in actuality it is more accurately portrayed as creative, somewhat improvised behaviour. It is therefore important to determine if the key issues suggested by the normative view of requirements engineering corresponds with the approaches being used in the real world. This paper outlines the theories of "rationalism" and "improvisation", and describes the major pitfalls of each. It then discusses how these pitfalls were encountered in the case study. Interestingly, although the methodology in this case study was very specific and firmly based in the "rational" paradigm, the users often improvised by privately making decisions on how certain aspects should or should not be implemented. As such, elements of both rationalism and improvisation were experienced. The paper concludes with the assertion that even where methodologies are developed and revised based on actual experiences in practice, as opposed to academic theory, it is not possible to devise an approach that is wholly methodical because amethodical aspects will always creep in during implementation.