Examination of customer satisfaction surveys in choice modelling to support engineering design

Lin He, Christopher Hoyle, Wei Chen*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


With the rapid expansion of customer satisfaction survey (CSS) data collected after product purchases, there is a growing interest in utilising such data for consumer choice modelling to guide engineering design. The question remains whether individual ratings can be directly used as a measure for qualitative product attributes in consumer choice modelling.A close examination of CSS data with respect to its applicability to consumer choice modelling is first provided in this work. Several key issues are identified, including ownership bias, differences in rating style, and missing choice alternatives' attributes. To alleviate these limitations, a systematic mixed logit-based choice modelling procedure is developed to incorporate the use of both quantitative and subjective rating measures in the model utility function, together with the consumer demographic attributes. The customer satisfaction index is introduced to measure consumer opinion in the form of ratings with respect to each design in the choice set. A case study using the real vehicle quality survey data acquired from J.D. Power and Associates demonstrates many of the key findings from this research including the characteristics of rating data as well as the applicability of the proposed integrated mixed-logit modelling procedure.

Original languageEnglish (US)
Pages (from-to)669-687
Number of pages19
JournalJournal of Engineering Design
Issue number10
StatePublished - Oct 1 2011


  • Consumer choice modelling
  • Consumer satisfaction index
  • Customer satisfaction survey
  • Engineering design
  • Mixed-logit model

ASJC Scopus subject areas

  • Engineering(all)

Fingerprint Dive into the research topics of 'Examination of customer satisfaction surveys in choice modelling to support engineering design'. Together they form a unique fingerprint.

Cite this