GENERALIZABILITY OF STUDENT-RATINGS OF UN IVERSITY TEACHING - CAN THEEVALUATION RESULTS OF ONE COURSE BE TRANSFERRED TO OTHER COURSES OF THE SAME UNIVERSITY TEACHER OR TO COURSES WITH EQUAL CONTENT BUT DIFFERENT TEACHERS

Results of student evaluation of university teaching can reflect the t eaching behaviour of an individual teacher, the composition of course participants or how interesting the subjects are. These factors can be analysed if ratings of several courses are compared: Comparisons of e valuation results of various courses of one teacher with results from courses on the same topic by other teachers enable us to draw conclusi ons on the personal or situational stability of evaluations. With the help of the Heidelberg Inventory on Teaching Evaluation (Heidelberger Inventar zur Lehrveranstaltungsevaluation, HILVE) the generalizability of student evaluations was analyzed in five investigations of four Ge rman Universities. Ita was shown that teacher variables (e.g. teaching competence) are rated in correspondence to the teacher, but student s cales or topic variables (e.g. active participation of topic interest) in correspondence to the respective course. Each dimension, however, measures to varying degrees the teaching behavior of the lecturer and aspects of the course. Thus, ratings of the teacher variables can be t raced back to teaching behaviour to a great extent but not entirely.