Minimum aberration designs for discrete choice experiments

A discrete choice experiment (DCE) is a survey method that gives insight into individual preferences for particular attributes. Traditionally, methods for constructing DCEs focus on identifying the individual effect of each attribute (a main effect). However, an interaction effect between two attrib...

Full description

Saved in:
Bibliographic Details
Published inJournal of statistical theory and practice Vol. 11; no. 2; pp. 339 - 360
Main Authors Jaynes, Jessica, Xu, Hongquan, Wong, Weng Kee
Format Journal Article
LanguageEnglish
Published Cham Taylor & Francis 03.04.2017
Springer International Publishing
Subjects
Aliasing
blocked design
estimation capacity
fractional factorial design
multinomial logit model
orthogonal array
Probability Theory and Stochastic Processes
Statistical Theory and Methods
Statistics
orthogonal array
62J15
multinomial logit model
62K15
Aliasing
estimation capacity
blocked design
fractional factorial design
62K10
Online AccessGet full text

Cover

More Information
Summary:A discrete choice experiment (DCE) is a survey method that gives insight into individual preferences for particular attributes. Traditionally, methods for constructing DCEs focus on identifying the individual effect of each attribute (a main effect). However, an interaction effect between two attributes (a two-factor interaction) better represents real-life trade-offs, and provides us a better understanding of subjects' competing preferences. In practice it is often unknown which two-factor interactions are significant. To address the uncertainty, we propose the use of minimum aberration blocked designs to construct DCEs. Such designs maximize the number of models with estimable two-factor interactions in a DCE with two-level attributes. We further extend the minimum aberration criteria to DCEs with mixed-level attributes and develop some general theoretical results.
ISSN:1559-8608
1559-8616
DOI:10.1080/15598608.2017.1299055