Are student evaluations of university teaching biased?

Questionnaires that use students as a source of information to evaluate university teaching are a common tool in university evaluation systems. The lecturers often question their value by alluding to the possibility that students may make biased judgments, linked to teaching traits or events not related to a fair assessment of the teaching activity. The main objective of this work is to examine the relationships between the characteristics of students and lecturers and the scores on the teaching evaluation questionnaire applied to students at the Complutense University of Madrid, in order to detect possible biased patterns in the evaluation they offer of their teachers. A hierarchical linear crossclassification model was used, with two levels, taking students as the first level and the lecturers as the second. The sample of this work is composed of 143,377 surveys, completed by 33,071 students, which involved the evaluation of 7,885 teaching activities and 3,922 university teachers in the academic year of 201617.The results show that the students' evaluations of their lecturers are mainly influenced by their interest in the subject, the age of the students and their lecturers and, to a lesser extent, attendance, hours of study and research quality. It should be noted that the type of undergraduate or master's degree studies, student's academic performance, and the lecturer's job category are not related to the teaching evaluations.After this analysis of the results, we cannot deduce the existence of invalidating biases derived from the evaluation of university teaching by questionnaires answered by the students