Models for Heterogeneous Variable Selection

Abstract

Marketing managers are interested in knowing how consumers will react to different product configurations. The product manager can change physical attributes through the design of the product and the perception of psychological attributes through promotion strategies. Because consumers are heterogeneous in their tastes and preferences, aggregate level estimates of attribute importance are insufficient to describe the market. New research methods focus on obtaining individual level estimates of attribute importance from a representative sample of consumers. Marketing researchers have procedural and statistical methods of obtaining measures of attribute importance for each respondent on each attribute. In laboratory or experimental choice settings, studies can be designed to help focus respondents' attention and processing of the product attributes. Bayesian methods of modeling heterogeneity shrink poorly measured individual level parameters to the overall or group level mean. However, it is erroneous to assume that consumers use all the product attributes in all brand choice situations. This thesis demonstrates that improved inference and predictive accuracy can be obtained by modeling which attributes are actually being used by consumers in different discrete choice situations. This thesis contributes new models for determining, at the individual level, which product attributes are being used by a consumer in a brand choice decision. The ii heterogeneous variable selection model extends current aggregate level models of Bayesian variable selection. This model assumes a distribution of heterogeneity with mass concentrated at 0 and away from 0 for each parameter. The pooled variable selection model allows the set of attributes used by an individual to vary by choice context. Examples of separate contexts include partial and full profile choice experiments or choice experiments and actual market place transactions. A hybrid model combines the heterogeneous and pooled variable selection models. The threshold variable selection model incorporates insights from an extended model of choice and provides a behavioral explanation of why certain product attributes are used. Tractable algorithms are introduced for estimating the proposed variable selection models. In the two empirical studies presented, a variable selection model fits the data better than baseline models with no variable selection and conventional distributions of heterogeneity. iii Dedicated to my wife, Teresa Minardi Gilbride iv ACKNOWLEDGMENTS

Skew and attribute non-attendance within the Bayesian mixed logit model

“…Although it is beneficial if such information is collected, constraints on survey size, or re-examination of prior datasets that did not ask such questions, may mean that it is not available. Several studies probabilistically allow for nonattendance without the need for self-reported indicators [3,10,13,21].…”

Modeling marketplace behavior

“…The posterior mean of γ i indicates the probability that a variable is significant in the model and is not zero. Gilbride et al (2006) propose an extension to George and McCulloch (1993) that allows variable selection to be present in models with heterogeneity, where different respondents are sensitive to different variables. The multivariate extension of the standard variable selection model is specified as:…”

“…Significant gains have been made in each of the areas mentioned above, including models of heterogeneity (Allenby and Rossi 1999), models for endogenous variables (Manchanda et al 2004) and models of variable selection (Gilbride et al 2006). Since research cannot simultaneously address all issues at once, these gains typically come in one area while using other, less advanced components in other areas of the model.…”

Monitoring Choice Task Attribute Attendance in Nonmarket Valuation of Multiple Park Management Services: Does It Matter?

“…Accounting for it tends to improve statistical fit of discrete choice models and allows analysts to run a sensitivity analysis of estimates of welfare measures to assumptions on attribute attendance. Some statistical models that endogenously account for the degree of serial AN-A have already been used successfully (Gilbride 2006, Rigby and Burton 2006, Scarpa et al 2009, Hess and Hensher 2009). However, the specific monitoring of choice task AN-A---as a form of intra-panel variation---is yet to be explored in the practice of environmental valuation, and it has only been minimally explored in other fields.…”