(Joint with Xia Wang, Department of Mathematical Sciences, University of Cincinnati and Joseph Pancras, Marketing Department, University of Connecticut)

Abstract: Investigating Nested Geographic Structure in Consumer Purchases: A Bayesian Dynamic Multi-scale Spatiotemporal Modeling Approach

Spatial modeling of consumer response data has gained increased interest recently in the marketing literature. One important dimension of spatial data is the determination of spatial scales for analysis, and the relevance of multi-scale spatial modeling to Marketing. In this presentation, we extend the (spatial) multi-scale model by incorporating both spatial and temporal dimensions.  We apply the recently developed dynamic multi-scale spatiotemporal modeling approach in the Bayesian statistics literature  to uncover spatial clusters of consumers who exhibit similar spatiotemporal behavior and gain insights on the market structure under hierarchical multi-scale levels. Our empirical application with a US company’s catalog purchase data for the period 1997-2001 reveals a nested geographic market structure that spans geopolitical boundaries such as state borders.  The multi-scale model also has better performance in estimation and prediction compared to several spatial and spatiotemporal models.  The model also provides statistical inference and can be easily implemented using scalable and computationally efficient Markov chain Monte Carlo (MCMC) method, underlining its potential for analyzing large spatiotemporal consumer purchase datasets.

Event Contact:
Tracie Hase   Phone:+1 (571) 858-3132