Breadth In Design Problem Scoping: Using Insights From Experts To Investigate Student Processes

Abstract

Because design plays a central role in engineering, it is important for engineering education programs to prepare students with design skills. By describing both novice and expert approaches to engineering design, researchers are contributing to the formulation of more specific design learning outcomes that may be addressed in curriculum design and program planning. One learning area where novices and experts differ is in how broadly they define engineering problems with which they are faced. This paper examines differences between how novices and experts approached the same hypothetical engineering problem. First-year students (n=124) and experienced engineers (n=4) were asked to identify factors they would take into account when designing a retaining wall system for the Mississippi River. Expert data were gathered using verbal protocol analysis, in which subjects were asked to "think aloud" as they addressed the retaining wall problem, and their statements were coded and interpreted. Novice data were gathered using a written protocol in which subjects were asked to simply list the factors on paper. Qualitative data were segmented into distinct ideas, which were then coded using a coding scheme with two dimensions of problem scoping breadth: physical location and frame of reference. We found that novices offered a greater proportion of factors from the natural and social frames of reference, versus technical and logistical frames, which indicated a rather broad approach the problem. We argue that this may reflect the novices' relative inexperience with engineering concepts. While the four experts' responses differed in terms of their representations through a "breadth of problem scoping" coding scheme, two of the responses echoed a characteristic top-down, breadth-first approach to design. The difference in protocols presents challenges in comparing expert and novice behavior, and methodological issues of collecting less information from a greater number of subjects versus collecting more information from fewer subjects were addressed. Because asking the experts to think aloud resulted in a rich data set, we employed narrative analysis to further investigate expert responses. The narrative analysis of expert problem scoping behavior suggested a sophisticated approach to situating problems and solutions in context. It highlighted several particular kinds of factors that the four experts in our sample were drawn to -existing engineered solutions, alternative design solutions, costs and benefits, priorities, and history. In addition, the narrative analysis illustrated the relationships between and among an expert's ideas, and what these relationships imply for the expert designers' thought processes.

Topics & Concepts