The use and misuse of student evaluations of teaching

Abstract

With the purpose of highlighting the validity and use of student evaluations of teaching (SETs), this article analyzes student grades and student evaluations of teaching performance along with nine other independent variables in the Spanish program of a major university. Data analyzed for this project represents four years of teaching and includes the evaluations and grades of 18,175 students. The most significant findings of the study are the following: There is a moderate correlation between low grades and low evaluations, but no correlation between high grades and high evaluations when all cases are considered together. Analysis and interpretation of the data suggests that SETs should not be used to compare language instructors. In addition, since the relationship between student evaluations and the actual merits of teaching performance has not been clearly identified, numerical values of those evaluations should not be used in critical personnel decisions such as retention, tenure and promotion of faculty, unless they are properly interpreted within a sound theory of teaching effectiveness.

Notes

1. The researcher is the undergraduate language program director in the Department of Spanish and Portuguese at San Diego State University, and oversees a large number of Spanish language courses. In some semesters, only the first three courses of Spanish had more than 100 sections all together.

2. The ‘grades’ variable does not include classes taught during the summer or classes taught abroad.

3. The highest correlation coefficient of this research project was the one found between age of the instructor and years of experience of the instructor (0.751). This is an obvious relationship that merits no further comment, as it was not part of the study.

4. By ‘first time’ the author refers to the first and/or second semester of GTAs’ teaching careers.

5. Data analysis is available upon request.