How Departments of Economics Evaluate Teaching

Abstract

Based on results from a 1999 national survey, William Becker and Michael Watts found that student evaluations of teaching were by far the most widely used, and often the only method used by economics departments, to evaluate teaching in undergraduate economics courses. To investigate whether departments of economics have moved beyond the use of student evaluations of teaching, in 2011 the current authors conducted a national survey of departments based largely on questions used in the 1999 survey. The surveys included items on how courses and teaching are evaluated, and on how that information is used in departmental promotion and salary decisions.

Keywords:

• evaluating teaching
• peer evaluation of teaching
• student evaluation of teaching

Keywords:

• JEL code
• A22

Notes

1. White (1995) reports findings on how departments assess teaching effectiveness. His results are from an informal survey to which 76 departments responded.

2. Despite the widespread popularity of SET across almost all schools and departments, and contrary to assertions by Marsh (2007) and some other supporters of SET, SET measures are only weakly correlated with other kinds of teaching evaluation measures, including peer evaluations, evaluations by trained observers, instructor self-evaluations, ratings from past students, and test scores or other measures of student performance. Becker (2000) notes the simple correlation coefficients between SET measures and the other measures cited in the literature are typically less than 0.7, and usually much lower, which means that student evaluation scores are associated with less than half the variation in other measures of teaching effectiveness. Weinberg, Fleisher, and Hashimoto (2009) show that end-of-term student evaluations are positively related to grades in a current class, but unrelated to students’ learning in later economics courses, after controlling for the grades in the earlier course.

3. The total mailed was 782, but four were sent to schools classified with associate degrees, two to schools with no Carnegie classification, and two to bad addresses. No responses were received from these schools.

4. The percentages for the location of department variables add up to more than 100 percent because some respondents marked more than one location.

5. The wording used on the surveys was “Electronically (e.g., via computers).”

6. The wording used on the surveys was, “In evaluating teaching as one component of decisions concerning annual raises, promotion, and tenure, how much weight is given to student evaluations of teaching? (e.g., if only student evaluations are used to evaluate teaching, the response would be 100%.) __________________________%”

7. Bosshardt and Watts (2001) found that grading rigor was an important component in determining the overall teaching ratings by students with instructors whose first language was not English.

8. A table showing the distribution of peer review results for the overall sample and broken down by Carnegie classifications is available on request from William Bosshardt.