Choosing Optimal Designs for Pair-Wise Metric Conjoint Experiments

Abstract

This paper presents an approach to the optimal fixed experimental

design for pair-wise metric conjoint studies. We first propose a linear

model suitable for metric paired comparison conjoint analysis.

Following Bayesian decision theory, optimal design problems in pairwise

metric conjoint analysis is then formally defined. Given the formal

definition of the experimental design problems, algorithms for the

identification for the optimal design are developed. The proposed

methodology is applied to a hypothetical conjoint experimental design

problem and major findings are discussed.