Considering usage context attributes in choice modeling has been shown to be important when product performance highly depends on the usage context. To build a reliable choice model, it is critical to first understand the relationship between usage context attributes and customer profile attributes, then to identify the market segmentation characterized by both sets of attributes, and finally to construct a choice model by integrating data from multiple sources. This is a complex procedure especially when a large number of customer attributes are potentially influential to the product choice. Using the hybrid electric vehicle (HEV) as an example, this paper presents a systematic procedure and the associated data analysis techniques for implementing each of the above steps. Usage context and customer profile attributes extracted from both National Household Travel Survey (NHTS) and Vehicle Quality Survey (VQS) data are first analyzed to understand the relationship between usage context attributes and customer profile attributes. Next the principal component analysis is utilized to identify the key characteristics of hybrid vehicle drivers, and to determine the market segmentations of HEV and the critical attributes to include in choice models. Before the two sets of data are combined for choice modeling, statistical analysis is used to test the compatibility of the two datasets. A pooled choice model created by incorporating usage context attributes illustrates the benefits of context-based choice modeling using data from multiple sources. Even though NHTS and VQS have been used in the literature to study transportation patterns and vehicle quality ratings, respectively, this work is the first to explore how they may be used together to benefit the study of customer preference for HEVs.

This content is only available via PDF.
You do not currently have access to this content.