Customer Loyalty Modeling under Logistic Function

Paul Louangrath,Paul I. Louangrath
DOI: https://doi.org/10.2139/ssrn.3011143
2017-01-01
SSRN Electronic Journal
Abstract:The objective of this paper is to present a case study of customer loyalty involving food services in a university canteen. The purpose is to provide a practical tool for modeling customer loyalty by using logistic function. Prior literature employed linear equation to model customer loyalty. This paper feels the gap in the literature by showing how logistic function could more accurately predict customer loyalty measurement. The data used in this paper came from 100 written surveys. The instrument employed non-Likert scale (0,1,2,3) to collect quantitative data. Nine variables were tested for their significant relationship to customer loyalty: service quality, customer satisfaction, trust, commitment, switching cost, corporate image, service recovery, emotion, and communication. McFadden R-squared for logistic modeling is 0.77 compared to the explanatory power under linear regression modeling, the ANOVA R-squared was 0.65. Logistic modeling has more explanatory power than linear regression. This study extends Kahneman-Tversky’s prospect theory from risk analysis to customer loyalty study.
What problem does this paper attempt to address?