Research on application of customer satisfaction index model — View_based on PLS and information entropy-weight method

Customer satisfaction index model belongs to the structural equation model (SEM). The modeling method of SEM mainly includes the linear structure relations based on the covariance structures (LISREL) and the path method based on partial least square (PLS). These two methods have different requirements for data. LISREL requires observed variables with normal distribution, while PLS has no strict restriction on data. So PLS is more suitable for modeling SEM. Entropy is the degree of irregular of the system. In SEM, the latent variable is the measured through observed variable. Less of information entropy of some measurement leads to lager of extent of variation and the provided information capacity, and more important role in latent variable in the measurement. The information entropy-weight method based on objective weighting is applied, which can avoid the subjective factors on the valuation of the observed variable. In this paper, the supermarket corporation is taken as an example; the path coefficient formation process of customer satisfaction index model is confirmed by means of the software of SPSS13.0 and SIMICA-P12.0. The results show that the scientific and practical modeling methods offer reference and consulting for customer satisfaction index model, and even for modeling SEM theoretically and practically.