This article discusses the various means to describe consumers' preference heterogeneity in choice models. It contrasts the use of continuous probability distributions to the use of discrete mixture models. Also, it presents estimation methods such as Markov Chain Monte Carlo methods and simulated likelihood. Finally, it opens areas for research such as the search for empirical generalizations, the testing of heterogeneity assumptions, and the separation of heterogeneity from state dependence and consideration sets. Link to the article
WEDEL, M., KAMAKURA, W., ARORA, N., BEMMAOR, A.C., CHIANG, J., ELROD, T. ... POULSEN, C.S. (1999). Discrete and Continuous Representations of Unobserved Heterogeneity in Choice Modeling. Marketing Letters, 10(3), pp. 219-232.