A latent structure double hurdle regression model for exploring heterogeneity in consumer search patterns

Heterogeneity in this process is accommodated vis a vis finite mixture distributions where distinct search groups or market segments can be identified via derived posterior probabilities of membership.

Abstract

According to a vast behavioral literature in Marketing, consumers typically engage in two sequential stages of information search in order to reduce uncertainty and maximize the anticipated benefits of a purchase. In the first stage, consumers initially retrieve product information stored in memory (internal search). If there is insufficient information in memory, the consumer then engages in external search for additional information from external stimuli in the market place. Consumers vary in the amount of each type of search they engage in depending upon their level of experience with the product category, socio-economic status, the importance of the purchase, product type, etc. This paper develops a latent structure methodology designed to model this sequential process. Here, the two sequential stages of search are explicitly modeled in a conditional, nested framework. In addition, heterogeneity in this process is accommodated vis á vis finite mixture distributions where distinct search groups or market segments can be identified via derived posterior probabilities of membership