CMStatistics 2017: Start Registration
View Submission - CMStatistics
B1399
Title: A review of optimal and efficient designs for choice experiments with partial profiles Authors:  Heiko Grossmann - Otto-von-Guericke-University Magdeburg (Germany) [presenting]
Abstract: Designs for choice experiments with many attributes require the respondents to process a large amount of information when the alternatives in the choice sets are specified by using all available attributes. The resulting complexity may prompt participants to use simplifying decision rules that violate the additivity assumption on the latent utility scale of the choice model which in turn may invalidate the statistical analysis. One approach to mitigating this problem is to use partial profiles which use only some of the attributes to specify the alternatives in each choice set. Recently, there has been a lot of interest in optimal and efficient designs for this type of choice experiment with contributions from different groups of authors. We present an overview of these developments including some of our own research. We focus on analytic results and corresponding design constructions which at the design stage make the so-called indifference assumption that the utility parameters are equal to zero.