Utility Models for Multiple Objective Decisions: Do They Accurately Represent Human Preferences?*

ABSTRACT Recent developments in multiattribute utility (MAU) theory in principle provide a formal basis for applying optimization methods to decision problems involving multiple competing objectives. In practice, MAU models and procedures are rarely applied. This paper is addressed to one major source of reluctance to employ these methods; namely, the belief that they cannot accurately reflect the structure of human preferences. This objection to MAU modeling procedures is inherently empirical. In fact, a large body of psychological research has been addressed to this question. This paper examines the relationship between formal MAU theory and the relevant psychological literature. A critical evaluation of this psychological research suggests that a variety of approaches to MAU modeling will typically yield valid mathematical representations of the preferences of individual decision makers.