Abstract
In this paper, the Bernoulli and the First Order brand switching models are put to test on two sets of panel data. Non-aggregate analysis indicates that a Bernoulli process may explain the brand choice behavior of consumers for frequently bought nondurable products. However, aggregate process seems to be of higher order. Moreover, the use of a simultaneous inference technique on consumer panel data is demonstrated.
Similar content being viewed by others
References
Anderson, T. W., and Goodman, L. A. (1957), “Statistical Inference About Markov Chains,”Annals of Mathematical Statistics, Vol. XXVIII, pp. 89–109.
Kesavan, R., and Srinivasan, V., (1976), “Bias in Binary Models of Brand Choice,” in Kenneth Bernhardt, Ed.,Marketing: 1776–1976 and Beyond, Proceedings of the 1976 Educators Conference of the American Marketing Association, pp. 115–119.
Lapin, L. L. (1973),Statistics for Modern Business Decisions, Harcourt Brace Jovanovich, Inc.
Lieberman, G., and Owen, D. (1961),Tables of Hypergeometric Distribution, Stanford University Press.
Massay, W. F., Montgomery, D. B., and Morrison, D. G. (1970),Stochastic Models of Buying Behavior, Cambridge, Mass.: M.I.T. Press.
Miller, R. G. (1966),Simultaneous Statistical Inference, McGraw-Hill.
Author information
Authors and Affiliations
Additional information
Based on a doctoral dissertation at the University of Rochester, Rochester, N.Y. 14627.
Rights and permissions
About this article
Cite this article
Kesavan, R. Empirical tests of two stochastic models of consumer brand choice. JAMS 8, 264–276 (1980). https://doi.org/10.1007/BF02721889
Issue Date:
DOI: https://doi.org/10.1007/BF02721889