Chapter 10
Discrete Choice Demand Modeling for Decision-Based Design


Decision-Based Design (DBD) is emerging as a rigorous approach to engineering design that recognizes the substantial role that decisions play in design and in other engineering activities, which are largely characterized by ambiguity, uncertainty, risk and trade-offs [1–6]. The DBD optimization seeks to maximize the utility of a designed artifact while considering the interests of both the producer and the end-users [1, 6]. Although there is great consensus that for a profit-driven company, the utility of a product should be a measure of the profit it brings, there is concern over using profit as the single criterion in DBD because of the belief that profit seems too difficult to model. One difficulty in modeling the profit is the construction of a reliable product demand model that is critical for assessing the revenue, the total product cost and eventually the profit.

In market research, there exist a number of approaches in demand modeling that explore the relationship between customer choice and product characteristics (attributes). Various analytical methods such as multiple discriminant analysis [7], factor analysis [8], multi-dimensional scaling [9], conjoint analysis [10–12] and Discrete Choice Analysis [32] have been used. They can be classified according to the type of data used (stated versus actual choice), type of model used (deterministic versus probabilistic) and the inclusion or noninclusion of customer heterogeneity. Even though demand modeling techniques exist in market research, they do not address the specific needs of engineering design, in particular for engineering decision-making.

Efforts have been made in the design community in recent years to extend the demand modeling techniques and incorporate customer preference information in product design [13–20]. Among them, the Comparing Multi-attribute Utility Values Approach from Li and Azarm [15] is a deterministic demand modeling approach, which estimates the demand by comparing deterministic multi-attribute utility values obtained through conjoint analysis. They also proposed a Customer Expected Utility Approach [16], which accounts for a range of attribute levels within which customers make purchase decisions and takes care of designers' preferences and uncertainty in achieving a desired attribute level. In recent years, the disaggregated probabilistic choice modeling approach in enterprise-driven engineering design applications has been employed [17–20]. Michalek et al. proposed a choice-based conjoint analysis approach within the multinomial logit (MNL) framework to analyze stated preference (SP) data. In this chapter, we illustrate how disaggregated probabilistic demand models based on discrete choice analysis (DCA) can be incorporated in a decision-based design (DBD) framework for making rational product design decisions.

  • Nomenclature
  • 10.1 Introduction
  • 10.2 Technical Background
  • 10.3 Implementing DCA for Demand Modeling in Engineering Design
  • 10.4 Walk-Through of a Typical MNL Model Estimation
  • 10.5 Industrial Example
  • 10.6 Conclusion
  • Acknowledgments
  • References
  • Problems
  • Appendix 10 Astatistical Goodness-of-Fit Measures

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In