Previous Table of Contents Next


5. OUR VALUES AND CONSIDERATIONS

We consider the computer system being designed as a future social actor in the organization (Stolterman, 1991). As Swedes, we have a culture of industrial democracy that demands participatory design or the “Scandinavian school” (Ehn, 1992; Floyd et al., 1989). We have also been highly influenced by thinking of interaction design as a process that begins at the very time a design assignment occurs and that the resulting system could be described as a social actor having a character (Stolterman, 1991). We do not believe there is ever an objective description of the real world. Modeling is therefore always an action where the modeler’s knowledge and values are a necessary component. Therefore, at Linné Data we created an object-oriented technique to create declarative enterprise models, which are easily understood and managed by “nonmodelers” in the organization (Ottersten and Göranson, 1993). The technique also makes it possible to capture all business rules without imperative expressions.

Generally, design evaluation suffers from the limitation that there is never one correct design. However, for interaction design, alternative ideas are relatively easy to evaluate. This is why we strongly emphasize the need for continuous evaluation and co-design. The evaluations are performed to ensure that the development process is kept on course. Our experience shows that during implementation some design ideas are transformed, resulting in unintentional changes in the design. This is not the result of “dumb” programmers, but rather an effect of their different roles (from designers) in the development process. Their goal is to produce (design) elegant code. When they are faced with a technical problem, they will find a work-around that sometimes reduces the usability of the system. Therefore, it is advantageous to have the external designer working part time during implementation in order to maintain the integrity of the original design.

Traditional, formal design methods too often end up being no more than modeling techniques or procedures that should or could be used. Furthermore, most formal methods do not explain the reasoning behind, or the values that lead to, the proposed actions. This results in methods that are even more difficult to use. Among system developers, in general, there is a widespread and justified mistrust of design methodologies (Stolterman, 1991). At the same time, some designers are seeking the method that will solve all problems. Our view is more pragmatic. The most important thing for us is not that the method is complete or 100% accurate but that it involves the users, and supports the designer in the design process. In fact, we strongly believe that there is no single method that can address all design questions.

External design could not be done without the co-operation of users. An experienced designer can usually generate desirable alternatives, but they are of little value until they are evaluated by users in the appropriate circumstances. This is a fact rarely understood by software designers and project leaders. Even customers tend to be insensitive to user issues. At LinnéData, we conducted a survey to assess the level of knowledge and interest in usability work among potential customers. Results showed that the major reason the customers have little interest in usability work is that they rarely, if ever, have the organizational responsibility for the people who are going to use the system.

There exist many techniques for generating a design. However, interaction design is more than just creating an artifact that appeals users. Interaction design is the very core of external design (i.e., producing an altered reality for people and contributing to the realization of an organization’s goals). The difficult part in interaction design is focusing on human needs. It is the myriad of human considerations that separates interaction design and software design, in general. However, we do not agree with those that claim interaction design is an art form. On the other hand, it seems obvious that some aspects of interaction design could never be formalized.

We emphasize that design is a continuous learning experience. In a specific project, the designer learns more and more about the design space, making the design process iterative. Over time, a designer’s knowledge is organized into larger chunks that consist of complex design issues and potential designs. The kind of learning that best describes interaction design is entertainment (i.e., the need of first-person-feeling and the urge to avoid interruption from the outside world or from thoughts that hinder the process) (Norman, 1993). This is consistent with our experience that interaction design is a social activity and is best done by people working in pairs (Winograd, 1996).

A computer system inevitably reflects a mental model of the way the users are believed to perceive the system’s structure. The design should be carefully crafted to reflect the users’ actual mental model. We use the term computer “interface” to emphasize the perspective of the human beings, not computers (Grudin, 1993).


Previous Table of Contents Next