Could not download file: This paper is available to authorised users only.


D. Keyson

Delft University of Technology (NETHERLANDS)
Balancing learning by doing with formal research
Researches through design focuses on the role of the product prototype as an instrument of design knowledge enquiry. The prototype can evolve in degrees of granularity, from interactive mockups to fully functional prototypes, as a means to formulate, develop and validate design knowledge. The design researcher can begin to explore complex product interaction issues in a realistic user context and reflect back on the design process and decisions made based on actual user interaction with the test prototype. Observations of how the prototype was experienced may be used to guide research through design as an iterative process, helping to evolve the product prototype. Given the design challenge of developing products that may be complex in terms of function and context of use, research through design clearly provides a means to demonstrate external validity of a design approach. However, though the research product prototype may develop into a high desirable design through a series of rapid design and validate iterations, it is often difficult for the designer to reverse engineer the resulting final prototype and form a substantial contribution to design knowledge in terms of generalizable results.

On the other hand controlled lab studies exploring a certain aspect of interaction, may lead to high internal validity but findings may be difficult to generalize and apply in context when not integrated as part of the total product experience. A middle ground is proposed in this paper based upon ten years of research through design projects in which controlled empirical studies were integrated into working prototypes so as to gain deeper insights into interaction issues and to contribute to design science. The central approach of this method referred herein to as empirical research through design is to create experimental variability in the product prototype so as to test the design question at hand and in context. The prototype variability has to be carefully defined so as not to confound the research question at hand. For example, if the design researcher is interested in how speech versus pen input might influence interaction on a tablet PC, than the visual interface should appear identical in the speech and pen input conditions as well as operation sequences. The design skills and creativity of the designer are critical in developing a balanced test platform. For example, users who may prefer speech input should not be steered towards pen input because of a poor speech input design. Secondly, the designer should be well informed on formal methods of statistics and research design techniques as common to the social sciences. In terms of small subject numbers, a common phenomenon in design studies involving physical prototypes; the researcher should strive to place a working prototype in context and test over time so as to increase the number of observations. Secondly, the designer should collect data in an objective manner so as not to bias results. Following a description of the empirical research through design process, several case studies, involving a host of student design projects, and the methods used to analyze the data are discussed.