Requirement Engineering: Ethnography and prototyping

Requirement Engineering

Introduction

The disciplines of Computer-Supported Cooperative Work (CSCW) and human-computer interaction (HCI) are both concerned with improving the systems design process so that systems better support the needs of users. In CSCW in particular, this has involved a turn to the social sciences for insights into how people work together, how work is socially organized, and what implications this might have for the design of systems to be operated by several people working together. The group of social scientists most involved in this is ethnomethodologists, who have undertaken a number of studies examining the nature of cooperative work supported by technology (Harper, Lamming, & Newman, 1992; Heath & Luff, 1992; Hughes, Randall, & Shapiro, 1992). However, the enthusiasm for the results of these studies has been balanced by a frustration with accessing and making use of the results in design.

Ethnographers

Ethnographers have started to examine the nature of the communication problem that exists between systems designers and ethnographers. The nature of this work has largely been a rethinking of how the results of ethnographers' fieldwork can be presented in such a way as to be useful to system designers (Hughes, O'Brien, Rodden, & Rouncefield, 1997; Hughes, O'Brien, Rodden, Rouncefield, & Sommerville, 1995). The work reported in this article approaches the same problem from an alternative direction by examining how an established system design notation can be used to represent the kinds of information forthcoming from ethnographic studies. Our approach was motivated by the need to link with the working practices of system developers and with the notations and tools that they use.

Prototyping

Prototyping is a rich area of study for human-computer interaction (HCI). It collapses the usual disjointing in space and time between developer and user. Although prototyping and user involvement are now mainstays of user-centered development approaches (Gould & Lewis, 1985), the implications leave many questions unanswered. Managing the practical problems of user involvement and prototyping is a very topical concern for commercial developers.

When talking with developers, unfavorable references sometimes would be made to nonprototyping approaches, in which various means might be employed to manipulate users into "signing off" on systems, with little understanding. Of course, it has to be understood that these utterances play a performative role in the contrast structure they set up. From one viewpoint, the strategies described here could be viewed simply as another manipulation of user expectations by developers. There is an agenda implicit in these methods of enlisting user participation in a process that is anticipated will result in a higher probability of acceptance than a "lobbing it over the wall" approach. However, to view this as simply manipulation by developers would be to miss the point. The developer (in Section 5.1), who reflected whether JRP sessions might be a kind of "trick" in that users are made to feel part of a team, went on to suggest a broader perspective for viewing this activity ("It's not. It's so true."). They take on a few responsibilities for the future ...