Basic Statistical Analysis and Modelling of Evaluation Data for Teaching

This thesis proposes a novel numerical scoring system, which efficiently evaluates the teaching effectiveness of the lecturers. Based upon the scores given in the student evaluation of teaching (SET), this numerical scoring system employes the factor score of one-factor model of data and yields the instructor rankings result as output. The other purpose of this paper is to discover determinants of SET scores, especially to examine whether factors which are normatively irrelevant to teaching quality matter or not. Results indicate that communication skill of lecturer & students’ reaction, course attributes and quality of lecture notes are three most significant factors which determines the student response to ”general overall ratings” of the course. The study suggests that class size and class meeting time also have some influence on that.