The results of a longitudinal study of groupware technology carried out over a 2 year period within a single organization is presented. The results of ordinal data derived from a questionnaire employed to determine user judgements of the usefulness of the technology for business task are analysed using a novel "best hypothesis" approach. This treatment uses formulations based upon AI and fuzzy mathematics and correspondence analysis.

Network software systems and groupware within organizations differ from other information technologies, requiring individuals to 'design' their own use. Users and groups can choose how to engage with these systems (Hassall, 1998), and use is dependent upon existing technological framing (Orlikowski, 1992). Groupware provides opportunities to study interaction between technological and organizational potentials. The action and structure duality of structuration theory (Giddens, 1984) points to the need for systemic understandings. Moreover, deconstructive schemes (e.g. Dudley and Hassall 1995,1996) demonstrate a plurality of overt and ulterior motivations in use. The Systemic Effector Model has been developed based upon longitudinal research in groupware implementation. This abstracted perspective relates choice of facility and design of action to important motivators at the individual and systemic levels. The genesis and explanatory power of the model is explored through survey and case study data.

The paper examines the role of decision support, or “choice making” systems and models based upon three groupings of ideas (or frameworks). Firstly, the philosophy of choice is examined with reference to the viewpoints of classicism, modernism and post-modernism as they relate to the way in which preferences are determined and valorised. Secondly, this tripartite framework is examined with reference to the philosophical works of Soren Kierkegaard who is sometimes regarded as the first existentialist philosopher. Third, some parallels are drawn between the models and frameworks thus far described and the psychotherapeutic model developed (initially) by Eric Berne, known as Transactional Analysis. Finally, a review and synthesis of some of the ideas introduced is attempted.

Two methods of analysing interval/ordinal questionnaire data based upon principles from fuzzy mathematics and artificial intelligence are described.