An interdisciplinary team of faculty and students from Art and Design, Business, Mathematics, Mechanical Engineering, and Psychology is exploring how to bridge the disciplinary viewpoints towards artifact creation and design. The emphasis is on developing a commonly understood quantification that employs modeling tools from the various disciplines, while also understanding the boundaries of rational quantification.

Design for construction of preference, Reactive and proactive design for preference construction
by Erin MacDonald, PhD Candidate

The premise that designers cannot "find" preferences or needs in a pre-existing state within customers challenges designers' skills, but incorporating the construction of preferences and thus inconsistent preferences into design methodologies will strengthen the success of products. Some preferences are prone to more inconsistency across decision context than others. Preference elicitation procedures, such as surveys, interviews, and observation, are useful in determining how best to identify and measure inconsistencies. Product design can accommodate inconsistent preference reactively, proactively, or with combination of the two approaches. When a designer or company takes a reactive approach to preference inconsistency, they design products that customers prefer across a variety of possible preference constructions - a type of robust design for uncertainty. In a proactive approach, designers can tailor product attributes to trigger specific preference constructions using product heuristics indentified in Research Theme 2. A proactive approach can carve out a new market segment that is particularly attractive to customers with inconsistent preferences. These customers may also be viewed as open to product innovations.

I have created a framework of preference inconsistencies that can be incorporated into product design methodologies. I am investigating product design optimization scenarios for a heterogeneous market, segmented using latent class analysis of discrete choice survey data. One segment's preference parameter, or utility coefficient, for recycled paper is highly inconsistent across preference measurements (assumed congruous to inconsistent preference for green products). In order to design an optimal product that incorporates marketing and engineering variables, this inconsistent utility coefficient must be addressed in the optimization framework. In the reactive design approach, a robust optimization is performed with the inconsistent utility coefficient modeled as an uncertain parameter. In the active design approach, the utility coefficient is instead modeled as a variable, with the assumption that appropriate product design and heuristics can "trigger" a specific value for the utility coefficient, and thus trigger one particular construction of preference for green products.

Downloadable materials: ODE poster