- Repository URL:
- Analogy; Idea Generation; Innovation; Design; Cognitive models
Idea generation and design-by-analogy are core parts of design. Designers need tools to assist them in developing creative and innovative ideas. Analogy is one such tool that helps designers solve design problems. It is a stimulus that helps generate innovative solutions to a design problem. It is used to generate novel ideas by transferring information (i.e. mapping elements) from a known domain (base) to an unknown domain (target). Multiple solutions can be developed based on a single analog and designers derive principles of design from the analogs (products) they experience. There is little research that discusses creating multiple solutions from a single analog or how multiple analogs can assist designers in mapping high level principles of design. Multiple paths are available to improve the design-by-analogy process and help designers understand the process better. This thesis explores two phases of design-by-analogy in which designers have difficulty generating multiple inferences from a single source analog and identifying high level principles given multiple example analogs in the presence of noise. Two hypotheses are proposed to explore the importance of analogies in design. 1. A lone designer is able to generate multiple inferences from a single source analog when instructed to do so. 2. The mapping of high level principles increases with the increase in the number of example analogs and decreases with the amount of noise. Two experiments, "Multiple Solutions" and "Multiple Analogies" are conducted to answer the proposed research questions and to understand how designers can become better analogical reasoners. The results from the "Multiple Solutions" experiment show that engineers, when directed to, can create multiple solutions from a single analog. Results from the "Multiple Analogies" experiment also satisfy the hypothesis that the mapping of high level principles increases with an increase in the number of analogs and decreases with distracters. A significant interaction is also observed between these two factors. The results indicate more future work with a greater sample size.