Determinants of students’ evaluation teaching according to their performance: an approach based on the relativeimportance metric

Student Evaluation of Teaching (SET) is an important tool to monitor teaching quality. In Italy, the SET is performed through the analysis of students’ opinion who fill out a questionnaire including a set of items related the teacher’s characteristics, the logistics, the organization of courses and the overall satisfaction on teaching quality. In this paper, we want to give simple statistical tools to construct an indicator of teaching quality according to student’s performance and to more important items which can influenced the student’s satisfaction. To build an indicator of student’s performance, we considered two variables, age and UEC (University Educational Credits) that influence significantly the student performance. Combining these two variables we obtained the ISP (Indicator of Student Performance). To investigate which items are more important, we use a relative importance metric based on Proportional Marginal Variance Decomposition (PMVD). PMVD metric allow us to overcome the problem of decomposition of variance analysis in an empirical study where the covariates are correlated.