In order to investigate the objectives of this study and test the hypotheses, LISREL 8.52 was employed to help us analyse the collected data. Five main statistical analyses were described as followed.
(1) Descriptive Statistical Analysis: in order to summarise the characteristics of respondents and better understand each research variable, descriptive statistical analysis was used to illustrate the means, and standard deviations of each research variable.
(2) Confirmatory Factor Analysis: CFA is used to test the reliability and validity. A reliability test for each construct was applied to assess the internal consistency from composite reliability. The composite reliability must be larger than 0.7 (Hair et al., 1998). Convergent validity can be assessed by the average extracted variances and were all above the recommended 0.5 level, and it can also be assessed by factor loading which should be greater than 0.5 (Hair et al., 1998).
(3) Structural Equation Modelling: SEM is used to verify the goodness of fit of the research framework and to describe the relationships among the construct variables. The LISREL 8.52 software package is used to analyse the relationships in the entire research model to find out the relationships among variables in this model.
(4) Fisher’s Z-transformation and ANOVA analysis: used to test the moderating effect of different customer perceived value on the relationship between customer satisfaction and customer loyalty.