Cognitive processes of similarity and combination in conceptual product design engineering

Student thesis: Doctoral Thesis


To create new products that satisfy human needs, product design engineers use their technical, manufacturing and creative knowledge to create candidate ideas for new products and develop them into final designs that can be manufactured. All new products have some basis in prior knowledge, and so design can be viewed as a process of knowledge recombination. A variety of methods and tools have been developed to help designers produce novel, useful design concepts through combinational thinking. One way to improve these design aids is by understanding the cognitive processes involved and tailoring methods and tools to foster effective cognitive processing and overcome cognitive constraints. Yet, despite the broad acknowledgement that designers do combine ideas to create new ones, little is known about how designers combine ideas to create new ones. In particular, there is no knowledge about how designers combine design concepts, which are candidate ideas produced earlier in the design process. The research presented in this thesis was conducted to model the cognitive processes involved in design concept combination and design concept similarity judgements. A deductive research approach was used to propose and test two cognitive models. The Dual-Process model of linguistic conceptual combination (Wisniewski, 1997a) was used as a basis for a cognitive model of design concept combination, and the dual-process view of similarity judgements was used as the basis of a model of design concept similarity judgements. Both models involve the same dual processes of comparison and scenario creation, and both models propose that the comparison process involves a process of alignment of structured mental representations. A series of research questions and hypotheses were proposed to test the models and a quasi-experimental research design was developed to evaluate them. The proposed Dual-Process model of design concept similarity judgements was tested in two experiments and it was concluded that student designers make similarity judgements of pairs of early-stage, sketch-based design concepts via a single process of comparison. In the first experiment (n=11), designers were asked to rate the similarity of pairs of design concepts and provide written explanations for their numerical ratings. The responses overwhelmingly indicated that designers make similarity judgements by focusing on the common and different features of the pair, i.e., a comparison process. In a second experiment (n=35), five predictions of the Structural Alignment model of similarity judgements were tested. It was found that similarity can be predicted as a function of the common and different features of a pair of design concepts, consistent with a comparison-based model of similarity judgements. However, only four of the five predictions were supported and so the Structural Alignment model was rejected. This means that it was not possible to draw conclusions about how the comparison process occurs. The proposed Dual-Process model of design concept combination was tested in one experiment (n=30). Student designers combined pairs of early-stage, sketch-based design concepts to create new design concepts that addressed the same brief. The proportion of combination types and their relationship with the similarity of the base concepts were measured and compared with the proposed model. Three kinds of combination were produced : (i) featural, (ii) relational and (iii) ambiguous. As the relative similarity of a pair of design concepts increases, the participants were increasingly likely to produce featural combinations and less likely to produce relational combinations. There was also evidence of a stimulus compatibility effect, a cut-off of relational combinations, and a defaulting to featural combinations. The featural and relational combinations and their relationship with similarity were consistent with the proposed model. However, the combination types were not fully accounted for. Thus, the proposed Dual-Process model does not fully capture the cognitive processes involved in design concept combination. Overall, the initial proposal that both similarity judgments and combination of design concepts occur via the same cognitive processes was incorrect. Comparison is involved in similarity judgements and may plausibly be involved in combination, but while there is evidence of a scenario creation process in design concept combination, there is none for design concept similarity judgements. Additional hypotheses and experiments are proposed to facilitate further research into the cognitive basis of the comparison processes in both models. The research and findings were critiqued to identify the advantages, disadvantages and opportunities and recommendations for future research.
Date of Award6 Mar 2023
Original languageEnglish
Awarding Institution
  • University Of Strathclyde
SponsorsEPSRC (Engineering and Physical Sciences Research Council)
SupervisorPaul Rodgers (Supervisor) & Madeleine Grealy (Supervisor)

Cite this