Studies On Customer Satisfaction Indices Model And Its Evaluation Methods

Nowadays, linear structural relationship and partial least squares regression are usually used to construct customer satisfaction indices models, and the two methods are discussed and studied deeply in this doctoral dissertation. Two customer satisfaction indices models for the evaluation of product and service are built, respectively, and demonstration analyses are made. And the methods for parameters estimation in linear structural equation model are discussed and the Sas procedure is given to compute the parameters. The theory, the steps for building the model and the achievement of the calculation of partial least squares regression are introduced. Moreover, analytic hierarchy process are used to determine a priority vector in the evaluation of customer satisfaction indices, and the scales, the computation of a priority vector, group decision, the consistency test and adjustment of a judgment matrix in AHP are studied and improved. To study influence factors of customer complaint, binary choice Probit model and binary choice Logit model of customer complaint are built, and two numerical examples are given, respectively. The theories and applications of panel data model,χ~2-test of multinomial distribution, contingency table analysis and variance analysis are introduced in the evaluation of customer satisfaction indices, and demonstration analyses are made.