The research aimed to answer two questions: "Is there a sequence in the loyalty stage framework?" and "What is the best way, linear effects versus curvilinear effects, in order to explain repurchase behavior?". The theoretical basis was adopted from Oliver (1999) and Fisbein. Oliver (1999) commented that an evolution in the loyalty theory could happen if researchers consider the theoretical framework described as cognition-affect-conation-action. This structure first theorizes that consumers become loyal in a cognitive sense, leading them to an emotional sense and then to a conative and, ultimately, an action, in order for buying new products. The methodology was the quantitative-descriptive survey, which is configured as a single cross-sectional survey. The findings indicated that in each loyalty stage that consumer passed (e.g., cognitive and conative), except for the affect, loyalty action had a significant increase in explanation coefficient and the best way to explain loyalty is through linear associations (vs. curvilinear). The practical implications suggest that the effects of traditional loyalty, with moderate levels vs. high levels, are better to generate higher store visit. The paper original value includes an advanced mathematical model to treat linear variables.
Loyalty; Dimensionality; Curvilinear effects; Non curvilinear effects; Retail
