Andrew Ainslie

Dissecting the Random Component of Utility

We illustrate and discuss several general issues associated with the random component of utility, or more generally “unobserved variability”. We posit a general conceptual framework that suggests a variance components view as an appropriate structure for unobserved variability. This framework suggests that “unobserved heterogeneity” is only one component of unobserved variability; hence, a more general view is required. We review a considerable amount of empirical research that suggests that random components are unlikely to be independent of systematic components, and random component variances are unlikely to be constant between or within individuals, time periods, locations, etc. We also review evidence that random components are functions of (elements of) systematic components. The latter suggests considerable caution in the use and interpretation of complex choice model specifications, in particular recently introduced forms of random parameter models that purport to estimate distributions of preference parameters. Several areas for future research are identified and discussed.

An Empirical Comparison of Logit Choice Models with Discrete Versus Continuous Representations of Heterogeneity

Currently, there is an important debate about the relative merits of models with discrete and continuous representations of consumer heterogeneity. In a recent JMR study, Andrews, Ansari, and Currim (2002; hereafter AAC) compared metric conjoint analysis models with discrete and continuous representations of heterogeneity and found no differences between the two models with respect to parameter recovery and prediction of ratings for holdout profiles. Models with continuous representations of heterogeneity fit the data better than models with discrete representations of heterogeneity. The goal of the current study is to compare the relative performance of logit choice models with discrete versus continuous representations of heterogeneity in terms of the accuracy of household-level parameters, fit, and forecasting accuracy. To accomplish this goal, the authors conduct an extensive simulation experiment with logit models in a scanner data context, using an experimental design based on AAC and other recent simulation studies. One of the main findings is that models with continuous and discrete representations of heterogeneity recover household-level parameter estimates and predict holdout choices about equally well except when the number of purchases per household is small, in which case the models with continuous representations perform very poorly. As in the AAC study, models with continuous representations of heterogeneity fit the data better.

UniBank and the analysis of the ExcursionCard customer database: A practical application of statistical techniques in database marketing

Abstract A common dilemma in direct marketing is that of which customers to target, and which to ignore. Once a profile has been gathered of a “good” or “ideal” customer, can those characteristics be identified in larger national databases? This article describes efforts to achieve this on behalf of a client, UniBank, who wishes to enlarge the market for a successful existing product, ExcursionCard. A large sample of the general population is compared with the existing ExcursionCard database in order to determine a customer profile with regard to responsiveness, profitability and risk. We demonstrate how a judicious combination of multivariate techniques and logistic regression can be used to help in determining whether these segments exist, what determines membership of the segments, and how to predict probable segment membership for potential clients.