Physical Models and Design Thinking: A Study of Functionality, Novelty and Variety of Ideas

Abstract

Engineering idea generation is a crucial part of new product development, and physical modeling is a widely used tool. Despite the physical models’ popularity in the idea generation process, little is known about their effects on design cognition. The existing literature provides contradicting guidelines about their use in the design process. Product design firms call for the frequent use of physical models, but some studies suggest that physical models induce design fixation. The psychological literature indicates that physical representations, by supporting designers’ mental models of physical phenomena, might lead to more feasible designs. The advantages and disadvantages of physical models as idea generation tools need to be clarified to help designers decide when and where to implement them. Based on prior studies and anecdotal evidence, two hypotheses are tested: (1) physical models supplement designer’s mental models and (2) physical models induce design fixation. Two between-subject idea generation experiments with novice designers are conducted to evaluate these hypotheses. In the first pilot experiment, the participants generate ideas in three conditions: sketching only, building, and building and testing. This study is followed by a second experiment, in which a new condition called constrained sketching is added. In each condition, participants use the representation implied by the name of the condition. The percentage of ideas satisfying all design requirements indicates the physical models’ effect on the designers’ mental models. Novelty and variety are used as metrics for design fixation. The percentage of functional ideas quantified shows significant variation across the sketching and building conditions, whereas novelty and variety show no differences. These results support the argument that physical models supplement novice designer’s mental models. No evidence of fixation is observed, which contradicts the results of the prior observational studies. Hypothesized reasons for the apparently contradictory results are also presented.