Evaluating quality of the didactics at university: The opportunities offered by latent class modeling

Purpose Students’ evaluation of teaching quality plays a major role in higher education. Satisfaction is not directly observable, nevertheless it can be measured through multi-item measurement scales. These instruments are extremely useful and their importance requires accurate development and validation procedures. The aim of this paper is showing how latent class analysis can improve the procedures for developing and validating a multi-item measurement scale for measuring students’ evaluation of teaching and, at the same time, provides a deeper insight in the phenomenon under investigation. Design/Methodology/Approach The traditional literature highlights specific protocols along with statistical instruments to be used for achieving this goal. However, these tools are suited for metric variables but they are adopted even when the nature of the observed variables is different, as it often occurs, since in many cases the items are ordinal. Latent class analysis takes explicitly into account the ordinal nature of the variables and also the fact that the object of interest is unobservable. Findings The data refer to the questionnaire to evaluate didactics to the students of the University of Padua. Within Latent class analysis allows an insight of scale properties, such as dimensionality, validity and reliability. Moreover, results provide a deeper view in the way students use the scale to report satisfaction suggesting to revise the instrument according to the suggestion by the National Agency for University Evaluation. Originality The paper gives an original contribution on two sides. On the side of methods, it introduces a more accurate methodology for evaluating scales to measure students' satisfaction. On the side of applications, it provides important suggestions to the university management to improve the process of quality of the didactics evaluation.