EFFECTS OF EXAMPLES ON THE RESULTS OF A DESIGN ACTIVITY

The role of search processes and the knowledge used in the activity of designing has been receiving increasing attention in the areas of artificial intelligence, the study of design, and cognitive psychology. More recently, with the development of expert systems, considerable attention has been focused on the representation and retrieval of expert knowledge. What has not been addressed, however, is whether or not an expert represents, accesses and utilizes all knowledge equivalently. The design-fixation effect, where pictures of design instances presented as part of a design problem result in the reproduction, by student and practising designers, of aspects of the presented instance, indicates that certain types of knowledge are more privileged in the design process than others. The results of a preliminary experiment, which examined the role of pictures of different types of instances, verbal descriptions of those instances, and variations in the familiarity of instances in the fixation effect, are reported. Pictorial information was shown to have no effect if the instance was unfamiliar, and equally familiar pictures were found to produce both design fixation and increased variety in design. Verbal descriptions of two of the pictured designs also produced effects, although these were much reduced in comparison with those of the pictorial material that produced fixation effects. While preliminary, these results indicate a particularly important direction for research, the results of which could be important for artificial intelligence, the study of design, and the psychology of design processes.