Human-centered vs. user-centered approaches to information system design

Journal of Information Technology Theory and Application, 2003 by Gasson, Susan

These insights demonstrate that we need to view IS design as involving problemexploration jointly with problem closure. Taking this approach would allow us to continually examine decisions concerning the role of IT within an organizational "system" of work (Checkland and Holwell, 1998). The remaining sections of this paper examine "state of the art" approaches to IS design that purport to focus explicitly on humancenteredness, to examine the extent to which these approaches support a problem-centered focus.

DEVELOPMENTS IN HUMANCENTERED DESIGN APPROACHES

Participatory Design

The socio-technical perspective is most apparent in the literature analysis of prototyping and participatory design. This area of work explicitly attempts to deal with the "multiple worlds" problem discussed above. IS stakeholders are placed in a situation where they can negotiate their requirements of an IS around a design exemplar - a prototype IT system, or a prototype work-system. But the attempt to balance the two domains tends to focus more on one domain than the other. Whilst, for example, Mumford's work in ETHICS (Mumford and Weir, 1979; Mumford, 1983) attempts the joint satisfaction of both social and technical interests, it deals almost exclusively with the design of work systems. Technology is viewed as infinitely configurable to suit the organization of workgroups, with no account taken of constraints imposed by either technology design or its implementation. More recent work (Butler and Fitzgerald, 2001; Lehaney, Clarke, Kimberlee and Spencer-Matthews, 1999) examines the ways in which user participation in decisions concerning the use of information technologies affects the outcome, but focus on participation in business process redefinition. While this is essential, it is not sufficient. We have discussed how goals may be subverted by the technical systems design and implementation processes that follow business process redefinition.

Muller et al. (1993) list a variety of methods for participatory design, classified by the position of the activity in the development cycle and by "who participates with whom in what". The latter axis ranges from "designers participate in users' worlds" to "users directly participate in design activities". For participatory design to be participatory, userworlds must be effectively represented in the design. But, as discussed above, there is a wide disparity in user "worlds". Participatory development has more potential to be politically disruptive and contentious than traditional (non-participatory) forms of system development, because it involves a wide variety of interests, with differing objectives and perspectives on how organizational work and responsibilities should change (Howcroft and Wilson, 2003; Winograd, 1996). This situation is therefore managed carefully in practice. System stakeholders are selected for participation on the basis of political affiliations and compliance, rather than for their understanding of organizational systems support and information requirements. This constrains user choice and significantly affects the potential to achieve a human-centered system design (Howcroft and Wilson, 2003). Users often have little choice about whether to participate. Even when trained in system development methods, users and other nontechnical stakeholders often cannot participate on an equal basis with IT professionals (Howcroft and Wilson, 2003; Kirsch and Beath, 1996). User views are often inadequately represented because of cost constraints, or a lack of appreciation of the significance of users' perspectives (Cavaye, 1995). Howcroft and Wilson (2003) argue that the user choice is significantly constrained by organizational managers, who predetermine boundaries for the scope of the new system, and who select who will participate in systems development and to what extent.