Cognitive processes in engineering design

• Main Author: Ball, Linden John
• Published: University of Plymouth 1990
• Subjects: 620.0042029; Computer aided design systems
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.277402

The central aim of the current research programme was to gain an understanding of the cognitive processes involved in engineering design. Since little previous empirical research has investigated this domain, two major exploratory studies were undertaken here. Study One monitored seven final-year students tackling extended design projects. Diary and interview data were used to construct detailed design behaviour graphs that decomposed activities into structured representations reflecting the goals and subgoals that were pursued. Study Two involved individual observation (using video) of six professional engineers "thinking-aloud" as they tackled a small-scale design problem in a laboratory setting. A taxonomic scheme was developed to classify all verbal protocol units and other observable behaviours. In interpreting the data extensive use was made of theoretical concepts (e. g. schemas and mental models) deriving from current research on human problem solving and thinking. Evidence indicated that the engineers studied had many similar methods of working which could be described at a high level of abstraction in terms of a common "design schema". A central aspect of this schema was a problem reduction strategy which was used to break down complex design problems into more manageable subproblems. The data additionally revealed certain differences in design strategy between engineers' solution modelling activities and also showed up tendencies toward error and suboptimal performance. In this latter respect a particularly common tendency was for designers to "satisfice", that is to focus exclusively on initial solution concepts rather than comparing alternatives with the aim of optimising choices. The general implications of the present findings are discussed in relation to both the training of design skills and the development of intelligent computer systems to aid or automate the design process. A final, smaller scale of experimental study is also reported which investigated the possibility of improving design processes via subtle interventions aimed at imposing greater structure on design behaviours.