Bayesian Analysis of Hierarchical Effects

The idea of hierarchical, sequential, or intermediate effects has long been posited in textbooks and academic literature. Hierarchical effects occur when relationships among variables are mediated through other variables. Challenges in studying hierarchical effects in marketing include the large number of items present in most commercial studies and the presence of heterogeneous relationships among the variables. Existing approaches have dealt with the large number of variables by employing a factor structure representation of the data and have used standard mixture distributions for representing different response segments. In this paper, we propose a Bayesian model for the analysis of hierarchical data using the actual response items and incorporating heterogeneity that better reflects consumer stages in a decision process. Cross-sectional data from a national brand-tracking study are used to illustrate our model, where we find empirical support for a hierarchical relationship among media recall, brand beliefs, and intended actions. We find these effects to be insignificant when measured with standard models and aggregate analyses. The proposed model is useful for understanding the influence of variables that lead to intermediate as opposed to direct effects on brand choice.