Empirical Generalizations in the Modeling of Consumer Choice

Abstract

Are there general algebraic laws which describe how consumers make choices from sets of alternatives? In this paper we review the verdict of research which has sought to answer this question. We focus on the functional forms which have been found to best characterize three component processes of consumer choice: those of attribute valuation, attribute integration, and choice. Our central conclusion is that there exists support for three major generalizations about the form of consumer decision processes: (1) subjective attribute valuations are a nonlinear, reference-point dependent, function of the corresponding objective measure of product attributes; (2) the integration rule which best describes how these attribute valuations are integrated to form overall valuations is multiplicative-multilinear, characterizing an overweighting of negative attribute information; and (3) the choice rule which links overall valuations of an option to the likelihood that it is chosen from a set is a member of a family of functions which recognize the attributewise proximity of a considered alternative to others in the set. The evidence supporting these generalizations is reviewed, as well as their implications for future theoretical and applied work in consumer choice modeling.