A Bayesian network approach to juxtapose brand engagement and behaviors of substantive interest in e-services

Abstract

This study juxtaposes brand engagement and behaviors of substantive interest in the context of eservices and utilizes Bayesian networks for an exploratory, data-driven approach. Enlisting a behavioral focus, it investigates the item-to-item relationships between brand engagement, attitudes, and conspicuous behaviors both online and offline. Two Bayesian networks were constructed from samples of city website users (N = 1089) and nonusers (N = 475). Results confirm that the five groups of nodes included in our constructed Bayesian networks were well distinguished as separate constructs and interrelated, centered on two nodes of brand engagement. Results also indicate that two nodes of brand engagement, “sense of belonging” and “participating in activities”, induced positive attitudes and were associated with conspicuous behaviors. Overall, this study demonstrates the dynamic, item-to-item relationships between engagement and behaviors of substantive interest. For managers, it is important in practice to discover the particular variables that play central roles among associated variables.