Groupware development can be conceived as one particular branch of software systems development. Research into groupware development faces both methodological challenges as in classical software development, and socio-technical issues as identified in the computer-supported cooperative working (CSCW) literature. On the one hand, the paradigm needs to accommodate the changing context for use, facilitate the effective communication between developers and users, and maintain the conceptual integrity of the system. On the other hand, it has to deal with the dynamic nature of groups and the emerging work practices during the development process. As Grudin (1988) pointed out, the social and organisational aspects within the groupware development (and use) often lead groupware to fail. Individual differences make it unlikely that two groups are in reality identical. Consequently, groupware development should be seen and treated as an organic process, in which the groupware is grown by the owners with the aid of professional developers rather than constructed by professional developers alone. In this thesis, I propose a new conception of efficacious groupware development which draws on the ancient Chinese philosophical notion of shi (as interpreted by Jullien (1995)).
In this thesis, I argue that Empirical Modelling (EM) potentially offers a conceptual framework well-suited for efficacious groupware development. In the process, I propose a new conceptual framework for practising an EM approach in a groupware development context. EM is a body of principles and tools which embraces an experimental, interactive, and open approach towards systems development through the exploration of observation, dependency and agency. The proposed conceptual framework, known as GroupPIE, is based on the principles of EM. This is built upon previous research into EM, particularly Sun's Distributed Empirical Modelling (DEM). In contrast to DEM, this thesis focuses on the micro-level of collaborative modelling. In particular, it considers how EM might facilitate the collaboration amongst the modellers and interaction between the modellers and the evolving artifact which takes place in groupware development.
The thesis draws on various case studies from undergraduate projects and research projects which have practised an EM approach to collaborative modelling. The case studies suggest that the participants' knowledge of the situation co-evolves with the artifact under construction and that there is role-shifting behaviour through the collaborative modelling. Drawing on the case studies, this thesis argues that an EM approach to collaborative modelling potentially facilitates genuine participation. This challenges the accepted ideas about the role of participants (or actors) and the relationship between them in the groupware development process. It also suggests that EM potentially facilitates a notion of participatory development which is “more” human-centred. On this basis, I argue that EM is potentially better-suited for realizing the vision of efficacious groupware development.