Compare methods

	Nested Logit Brand Choice	Hierarchical Bayes Choice Model
Field	Marketing	Marketing
Family	Regression model	Regression model
Year of origin	1978	2005
Originator	Daniel McFadden	Peter E. Rossi, Greg M. Allenby & Robert McCulloch
Type	Generalized-extreme-value discrete-choice model	Hierarchical Bayesian random-coefficients discrete-choice model
Seminal source	McFadden, D. (1978). Modelling the Choice of Residential Location. In A. Karlqvist, L. Lundqvist, F. Snickars, & J. Weibull (Eds.), Spatial Interaction Theory and Planning Models (pp. 75-96). North-Holland. ISBN: 9780444851826	Rossi, P. E., Allenby, G. M., & McCulloch, R. (2005). Bayesian Statistics and Marketing. John Wiley & Sons. ISBN: 9780470863671
Aliases	Nested Multinomial Logit, Hierarchical Choice Model, Tree-Structured Logit, GEV Nested Logit	HB Choice Model, Bayesian Random-Coefficients Logit, Hierarchical Bayesian Conjoint, Individual-Level Partworth Model
Related	3	3
Summary	The nested logit model of brand choice relaxes the restrictive independence-of-irrelevant-alternatives (IIA) assumption of the standard multinomial logit by grouping similar alternatives into nests. Developed by Daniel McFadden as a member of the generalized-extreme-value (GEV) family, it allows the unobserved utilities of alternatives within the same nest to be correlated while keeping a tractable closed form. In a brand-choice setting the natural structure is a tree: consumers first effectively choose a category, sub-category, or product form and then a brand within it, with an inclusive-value term carrying the expected utility of the lower level up to the upper level. The dissimilarity parameter on each nest measures within-nest correlation and reduces to ordinary logit when it equals one. The result is a model whose substitution patterns are far more realistic than plain logit — a price cut on one brand draws disproportionately from its nest-mates — while remaining estimable by maximum likelihood. It is a workhorse for choice analysis when alternatives fall into obvious clusters.	Hierarchical Bayes (HB) choice models estimate a separate set of preference weights — partworths — for every individual respondent, while borrowing strength across respondents through a shared population distribution. The model has two levels: at the lower level each person's choices follow a logit driven by their own coefficients, and at the upper level those individual coefficients are treated as draws from a common multivariate distribution whose mean and covariance are themselves estimated. Inference is Bayesian and proceeds by Markov chain Monte Carlo — typically Gibbs sampling with Metropolis steps — which yields a full posterior for each respondent's partworths rather than a single point estimate. The approach, codified by Rossi, Allenby, and McCulloch, solved a long-standing problem in choice modeling: how to recover genuine individual-level heterogeneity from the sparse data each person provides. Sparse individual estimates are stabilized by shrinkage toward the population mean, giving reliable person-level coefficients usable for segmentation, targeting, and realistic market simulation. HB is now the default estimator for conjoint and scanner-based choice analysis.