The Use of Logistic Regression Analysis to Evaluate Customer Satisfaction with a Telephone Hotline

Abstract

This article describes a method used to evaluate customer satisfaction with a telephone Hotline that, at the time of the study, received more than 200,000 calls annually. The statistical analysis, which is based on responses to a customer survey, specifies three logistic regression models as a set of recursive equations estimated in a two-stage process. The first stage analysis estimates two equations designed to measure service quality characteristics that contribute to customer satisfaction. The second stage analysis estimates the likelihood that customers were satisfied with their Hotline experience, as a function of the service quality characteristics and other factors. The results of the analysis provided several practical areas lor improving customer satisfaction with the Hotline.