An Analysis of Courses Evaluation Through Clustering

Students of the University of Florence (Italy), before taking an exam, are required to assess different aspects related to the course organization and to the teaching. The data concerning the evaluation of the courses of the Computer Science Program from 2001/2002 to 2007/2008 academic years were collected and linked to the results of students: the grades obtained in the corresponding exams and the delays, with respect to the end of the courses, with which exams were taken. After this preprocessing phase, we used clustering techniques to analyze data and we highlighted a correlation between courses evaluation and the corresponding average student results, as well as regularities among groups of courses over the years. Our analysis can be used to detect possible improvements in the organization and teaching of the degree program and applied to any university context collecting similar data.