Efecto de los métodos de estimación en las modelaciones de estructuras de covarianzas sobre un modelo estructural de evaluación del servicio de clases

Abstract

The purpose of this research was to perform a comparative analysis of methods for estimating parameters of MEC, in order to determine the most appropriate method, specifically using a structural model of class service evaluation. This model consists of the relationships between perceived service quality and perceived value classes on behavioral intentions of postclass, acting as a mediator variable student satisfaction. The methodology for establishing the most appropriate method is based on the limiting factors of the method considered (assumption of normality and sample size) and aspects of the superiority of the model (predictive power and goodness of fit indices). It has been established that the best method for estimating parameters of an structural model for service class evaluation is theunweighted least squares (ULS), because: 1) it was found that the data did not meet the assumption of multivariate normality, 2) is the only method in which the model presents indices appropriate fit, and 3) the model has high predictive power. In general it has been observed that there are percentage deviations from the values of the paths of the model using the various methods compared with the ULS ranging from -45% to 47%. Using the ULS method has been established that student satisfaction is a powerful mediator between service quality and perceived value classes on behavioral intentions of postclass. Using this method, it was established that student satisfaction is a powerful mediator between class service quality and perceived value on after-class behavioral intentions.