Most TAs get one piece of feedback on their teaching: the end-of-term evaluations, which arrive after the students are gone and say little you can act on. You can do much better, at little cost. A short anonymous survey a few weeks in, a quick question at the end of a section, a colleague watching twenty minutes of your lab, a recording of yourself: each tells you something you can change while it still matters. It also gives you the evidence a teaching statement and portfolio need. This page covers each source of feedback, a survey you can use, how to ask for an observation, how to report back to students, and how to read end-of-term evaluations sensibly.
Defaults that work. If you only take four things from this page:
- Ask early: a short anonymous survey around weeks 3 to 6, while you can still change things.
- Tell students what you heard and what you'll change, within a week.
- Ask a colleague to watch for one or two specific things, not "how did I do?"
- Read end-of-term evaluations as one source among several, and look for patterns, not single scores.
Midsemester feedback
- Timing: weeks 3 to 6 of a semester. Not on the day you hand back graded exams or projects.
- Keep it short and anonymous: two to five questions, about ten minutes in section, or an online form. If you do it on paper in the room, step out while students write.
- Make it about your section or lab, not the course or the professor, and say so.
- Word it around learning: "What helps you learn?" gets more useful answers than "Do you like section?"
A survey you can use
- On a scale of 1 to 5, how much does this section or lab help you make progress on the assignments? Why?
- What should your TA start doing to help you learn? (For example: live debugging, more worked examples, more time at your bench.)
- What should your TA stop doing, or do less of?
- What should your TA keep doing because it helps you learn?
- What is one topic, tool or error message you're still stuck on?
Reading the results
Sort comments into themes and count them. One sharp comment is memorable but may be one person; ten comments saying the same mild thing are the real message. Pick one or two changes you can actually make.
Reporting back (five minutes, next session)
- Thank them, and say how many responded.
- Share two or three themes, including what is going well.
- Name the one or two changes starting now.
- Name what can't change, and why: "The deadline is set by the course."
- Say how they can keep telling you things.
Closing the loop is what makes students take the next survey seriously.
Quick checks every session
A one-minute paper at the end of section ("What was the most important thing today? What's still unclear?") or a "muddiest point" question tells you what to start with next time. Read them the same day, and open the next section with the most common confusion.
Peer observation
- Ask a fellow TA or your instructor, and offer to return the favor.
- Choose one or two things to watch. A narrow focus gets useful notes; "how did I do?" gets "it was good."
- Meet briefly beforehand to share your plan, and debrief within two days.
- Ask for description, not judgment: what happened, when, how often. Observation protocols like COPUS record what students and the instructor are doing every two minutes, without rating anything.
Things an observer can watch for (pick one or two)
- Who talks and who doesn't; whether some students are never reached.
- How long you wait after asking a question.
- How time splits between you talking, students coding, and students working in pairs.
- When students ask for help, whether you give the answer or ask a question back.
- Whether you open with the goal for the session and close by summarizing it.
- Pacing during live coding: can students keep up?
A request you can send
"Hi ___, I'm working on my teaching in ___ and would value a second pair of eyes. Could you observe 30 to 50 minutes of my [day and time] lab sometime in the next two weeks? I'd like you to watch for one or two things: [from the list]. Could we meet for fifteen minutes beforehand, and for twenty minutes within two days after? Happy to do the same for you. Thanks!"
Recording yourself
Record one section, with permission and following your course's rules on recording students, or record just yourself. Watch it once with the same focus list an observer would use. Most people notice right away how little they wait after a question, or how fast they talk during a demo.
A consultation
Many teaching centers offer confidential consultations and small-group feedback sessions, in which a consultant talks with your students while you are out of the room and then summarizes what they said. Research on instructors finds that feedback helps more when someone helps you interpret it. Ask about two weeks in advance.
End-of-term evaluations
Read them, but read them carefully. Large studies have found that student evaluation scores are not reliably related to how much students learn, and that they show bias, for example against women instructors. They are still useful for comments about specific things you did. Look for themes across several terms, and put them next to your midsemester surveys, observations and your own notes. A teaching portfolio built from several sources is far more convincing than a single score.
Turning feedback into change
Stephen Brookfield suggests looking at your teaching through four lenses: your students' eyes, your colleagues', your own experience, and the research. Each catches what the others miss. When the sources agree, act. When one change seems promising, test it properly; Trying a Small Teaching Experiment shows how.
Further reading
Guides
- Yale Poorvu Center, Mid-Semester Student Feedback.
- Yale Poorvu Center, Teaching Consultations and Observations.
- Columbia CTL, Early and Mid-Semester Student Feedback: Start, Stop, Continue.
- CMU Eberly Center, Early Course Feedback.
- CMU Eberly Center, Using Classroom Assessment Techniques: the minute paper and muddiest point.
- UC Berkeley GSI Teaching & Resource Center, Conducting a Midterm Assessment: written for graduate TAs.
- Stephen Brookfield, Becoming a Critically Reflective Teacher.
Research
- Smith et al., "The Classroom Observation Protocol for Undergraduate STEM (COPUS)" (2013).
- Cohen, "Effectiveness of student-rating feedback for improving college instruction" (1980).
- Finelli et al., "Utilizing Instructional Consultations to Enhance the Teaching Performance of Engineering Faculty" (2008).
- Boring, Ottoboni and Stark, "Student evaluations of teaching (mostly) do not measure teaching effectiveness" (2016).
- Uttl, White and Gonzalez, "Meta-analysis of faculty's teaching effectiveness" (2017).
- Mitchell and Martin, "Gender Bias in Student Evaluations" (2018).
Related pages: Teaching Statements and Portfolios, Trying a Small Teaching Experiment and Covering a Lecture.