Goodness of fit measures for logistic regression model: an application for students’ evaluations of university teaching

This study investigates the assessment of university teachers. The evaluation of university teaching plays an important role in the university system in the light of the fact, that a part of the financial transfer from educational ministry is based on the quality of teaching. From Statistical point of view the use of regression models has been widely applied in this area. The logistic regression model is particularly used as discrete choice model using dichotomous dependent variable. For many regression analyses the lack of a goodness-of-fit measure is more important than coefficient interpretability. The goal of this paper is to present an overview of a few easily employed methods for assessing the model fitness of Logistic Regression Model by Pseudo-\(R^{2}\).Moreover the assessment is carried out through a simulation study to analyse the pattern (behaviour) of each measure, with precise focus on change of multiple correlation among the variables. In this paper a survey on student satisfaction (SS) of university teaching system was conducted. The instrument used in this paper is a questionnaire proposed by a different research group in 2010. The collected data was elaborated by a full reflective Structural Equation Model using PLS path model estimation. The initial results showed that the influence of the Organization and Infrastructures on the Student Satisfaction were statistically insignificant. Therefore a more complex model was supposed, the final results showed that the influence of Organization and Infrastructures on the SS was indirect, that is the Organization and the Infrastructures exert an influence upon the SS through the Didactics.