A Hybrid Utility Estimation Model for Conjoint Analysis

Abstract

Increasingly, appliers of conjoint analysis are being faced with the need to reduce data collection demands on respondents while still obtaining enough data to estimate individual utility functions. The authors propose a model that combines the ease of self-explicated utility measurement with the greater generality of decompositional models to develop estimated utility functions that maintain individual differences. The model is applied to a conjoint study involving physicians’ evaluations of a new antibiotic drug. The paper concludes with suggestions for possible extensions of the approach.