How do the relationships between variables impact behavioural loyalty? The Qualitative Multinomial Distribution (QMD) applied to wine consumption

Until nowadays, the studies on consumers? behavioural loyalty allowed the analysis of a brand, category or product attribute at a time. However, it has never been possible to observe and measure the relationships between brands/attributes from a loyalty perspective. Therefore, the paper aims at filling this gap presenting an innovative methodology, available through a new statistical distribution ? called Qualitative Multinomial Distribution (QMD) ? which detects the presence and the impact in terms of loyalty of interaction and correlation effects between brands categories and/or product attributes in both a partial and a comprehensive way. In this study, it has been shown that a comprehensive analysis offers a better representation of the connections between product attributes (or brand categories) than partial analyses. However, the QMD allows observing both approaches before deciding what the most suitable one is, thanks to a series of powerful diagnostic tools.