Representing Heterogeneity in Consumer Response Models 1996 Choice Conference Participants

Abstract

We define sources of heterogeneity in consumer utility functions related to individual differences in response tendencies, drivers of utility, form of the consumer utility function, perceptions of attributes, stated expendencies, and stochasticity. A variety of alternative modeling approaches are reviewed that accommodate subsets of these various sources including cluster-wise regression, latent structure models, compound distributions, random coefficients models, etc. We conclude by defining a number of promising research areas in this field.

Keywords: Heterogeneity, latent structure models, cluster-wise regression, random coefficients models, compound distributions