Modeling patronage behavior: a tri‐partite conceptualization

Purpose – By using mall patronage behavior as its context, this study aims to develop and test a patronage model consisting of three dimensions. Design/methodology/approach – Data for the study are collected through self‐administered questionnaires from residents in an MSA in the USA. Usable responses are obtained from 319 residents. Findings – Results suggest that mall patronage behavior can be represented as a global construct with three viable components (utilitarian, hedonic and accessibility). Research limitations/implications – The study is limited to a particular sample. Replications among other samples in the study locale and elsewhere are needed to validate the current findings. Practical implications – The results enable management to look at shoppers' patronage behaviors at three levels. At the individual attribute level (first level of abstraction), management may identify areas that need special attention. At the second level of abstraction (the latent construct), choice attributes can be combined into reliable and valid composite scores across dimensions and can reveal information that is not readily available by the individual attributes. At the third and highest level of abstraction, the one second‐order factor with three first‐order factors as its reflective indicators provides management with a single metric for comparing a mall with its competitors or other malls owned by the corporation. Originality/value – The model tested here explicitly recognizes accessibility as a distinct patronage dimension and expands the domain of inquiry beyond the initial attribute level to the first‐order and the second‐order composite levels.