Course Evaluation Design Playbook

Course evaluations are notoriously easy to get wrong - either too generic to act on, or so tied to how much students liked the instructor personally that the useful signal gets lost. A better structure separates the two.

1. Split instructor and course questions

"Was the instructor engaging?" and "did the course structure help you learn the material?" measure different things - a well-liked instructor can still run a poorly organized course, and vice versa. Ask both, separately.

2. Ask about specific course elements

Rate the syllabus clarity, workload pacing, and assignment usefulness individually rather than one overall "how was the course" score - specific ratings point to what to actually change next term.

3. Reduce popularity bias where you can

Timing matters: collecting evaluations right after a final exam, or right after grades post, skews results toward how students feel about their grade rather than the course itself. Collect before final grades are released.

4. Include one open-text question, not five

"What's one specific thing that would improve this course?" produces more usable feedback than several open fields that mostly repeat each other and drive down completion.

5. Report trends across terms, not one term in isolation

A single term's scores are noisy with a small class size. Track the same course over several terms to tell a real trend from normal variation.