Hybrid Models for Conjoint Analysis: An Expository Review

A classification of hybrid models is presented, followed by a review of their comparative performance in cross-validation tests, and Suggestions are offered on future studies that are essential before the role of hybrids in conjoint methods can be evaluated properly.

Abstract

Over the past few years hybrid models for conjoint analysis have been developed to reduce data collection effort and time. Hybrid models combine features of self-explicated utility measurement with more traditional conjoint analysis. A classification of hybrid models is presented, followed by a review of their comparative performance in cross-validation tests. Though hybrid models represent an attempt to cope with an important practical problem in industry applications of conjoint techniques, these models entail a number of untested assumptions requiring further theoretical analysis and empirical research. Suggestions are offered on future studies that are essential before the role of hybrid models in conjoint methods can be evaluated properly.