What Your Teaching Evaluation Scores Aren’t Telling You (and What to Do About It)

It’s the week after grades post, and you open the evaluation portal. You tell yourself you’re just going to skim the numbers. Apprehensively you scroll down to the comments, and there it is, one anonymous line, permanently attached to your academic record, that you will remember in vivid detail three years from now, long after you’ve forgotten the 11 students who wrote something kind.

If that paragraph felt uncomfortably familiar, you are not alone, and you are not doing anything wrong. Student evaluations of teaching (SET) are one of the strangest dossier documents we are asked to treat as professional evidence: They ask us to solicit feedback from the product of our teaching (students), not the customer (the employers and communities that inherit our graduates) of it. They are anonymous, emotionally loaded, and often written by 19-year-olds at 11:47 p.m. the night before the portal closes. The research on bias in SET is well known. What you need is a way to open that portal without dreading it and a system to reduce the emotional wear and tear of figuring out which comments deserve your attention.

This recommended process begins logistically, because how you read your evaluations matters almost as much as how you interpret them.

Read them all in one sitting. Do not let them trickle in across a week of stolen five-minute breaks between meetings. One hard hour beats five days of dread and drain.
Plan something to look forward to afterward. I read mine (in full) in the early afternoon because I need some time to process them before I pour myself an adult beverage that evening and call it well earned. Whatever your version of that is, treat the evening after evaluation-reading like the evening after a hard meeting, because it is one.
If something cuts deep, defang it before you analyze it. I have started pasting the harshest comments into an AI tool and asking it to roast the emotionally fueled comments back at me in the voice of Tina Fey and Amy Poehler doing a Golden Globes monologue. It sounds ridiculous. It works. It makes me chuckle and defuses those emotional grenades. This neutralization makes it easier to sort the comments pragmatically.

The Three-Bucket Sort

The Three-Bucket Sort is a framework I would encourage you to use along with a few strategic tools.

Importantly, distance this sorting process from the day you submit final grades. I would recommend at least 48 hours in between. This isn’t about being thick-skinned; it’s about not judging your teaching while the possible sting of the feedback is still being processed.

Bucket 1: The Pattern. A specific, describable behavior that shows up independently across multiple students, in their own words. “Lectures moved too fast,” “unclear what would be on the exam,” “group project instructions changed twice.” These should be viewed as constructive, concrete, and fixable.

Bucket 2: The Outlier. A single emotionally charged comment, often tied to a grade dispute or one bad classroom day. Nothing else in the set corroborates it. Worth a moment of reflection, not a course redesign.

Bucket 3: Static. Comments about your appearance, voice, accent, age, identity, or personality rather than your teaching, plus the zero-elaboration “boring” or “hard.” Static can sting; however, without a pattern it is often not actionable. Naming a comment as static and moving on is accurate categorization, not avoidance.

How many comments make a pattern? The answer: It scales with your response rate, not a fixed number.

Small to mid-sized class (under 75 respondents): Look for the issue appearing in roughly 10% or more of written comments, signaling a pattern.
Large lecture (75+ respondents): A raw count of 3–4 is often noise; look for the theme recurring across multiple sections or semesters of the same course before treating it as a signal.

As you review the comments, focus on one thought process: “Could you change this?”

“Class was at 8 a.m.” is a pattern of complaint but not something you control.
“I didn’t understand the grading rubric until after the first assignment was due” is a pattern that’s yours to fix. Sort for actionability, not just frequency.

Next, run it through one more filter before you act on it: “Whose learning does this actually serve?” Every pattern comment comes from somewhere in the room, and the room isn’t one level of student—it is the majority working at grade level, the students below the middle who need more scaffolding to catch up, and the students above the middle who are hungry to go further. A complaint about pace, for instance, might be entirely valid for the middle of your class and still be the wrong thing to fix if slowing down would shortchange the students working hardest to keep up, or bore the students above the fold.

The patterns worth acting on are the ones that help the middle of the room without abandoning either end, and pedagogically, they usually call for a differentiation fix (an optional scaffold, an extension problem) rather than a blanket one.

Catch It Before It’s Official: The Mid-Semester Check-In

A technique to soften your official evaluations is to ask for feedback while you can still act upon it. Cohen’s (1980) meta-analysis of 22 studies found that instructors who collected and acted on mid-semester feedback earned significantly higher end-of-semester ratings than those who didn’t, a gain equivalent to roughly 15 percentile points. Overall and Marsh (1979) traced the mechanism to timing: Feedback gathered while a course is still in motion changes both what an instructor does and how students perceive the rest of the term. A five-question, anonymous check-in does this well. Frame it as a genuine invitation to improve the remainder of the course.

Copy and paste this into a survey tool:

What is helping you learn in this course so far?
What is one thing that’s getting in the way of your learning?
Is the pace of the course too fast, too slow, or about right?
Is there anything about assignments, grading, or deadlines that’s unclear to you?
What’s one thing I could change for the rest of the semester?

Two things make this effective:

Close the loop. Spend time in your next class telling students what you heard and what you’re changing or why you’re not. This single step does more for your end-of-semester scores than anything else, because students who feel heard write kinder, more accurate comments in the final evaluation.
Only survey what you can act on mid-semester. Save deliberations like “was the textbook worth the cost” for your postmortem.

Framing SET in Your Portfolio: Don’t Let One Semester Write the Narrative

Every promotion and tenure process weighs SET differently. Having moved through the review processes at an R1 and a teaching-focused institution and now working at a third, AACSB-accredited college, I know a constant is this: Committees aren’t just looking at your numbers. They’re looking at whether you understand your numbers. A rough semester, unexplained, reads as a red flag. Contextualized, it reads as evidence of a reflective teacher who is willing to try new teaching approaches.

A few moves that consistently strengthen a teaching portfolio, regardless of institution type:

Show a trend line, not a snapshot. Committees are trained to notice a bad semester sitting in an otherwise solid multiyear trend. A simple line graph of scores by semester does more than paragraphs of explanation.
Name the outlier semester briefly, without over-apologizing. One or two sentences: what happened, what you changed. “In Spring 2023, I redesigned this course’s group project structure after evaluations flagged confusion; the following semester’s clarity scores rose from X to Y.” Don’t relitigate.
Triangulate. Pair SET scores with a peer or chair observation and any before/after materials from a course revision. A portfolio built on SET alone makes the committee interpret it for you; one with corroborating evidence narrates the story you want to tell.
Know your institution’s actual weighting. Ask your department chair or a recently tenured colleague how SET will be used in your review, and write to that expectation.

The Bottom Line

You cannot control what a student writes in an anonymous box at 11:47 p.m. You can control how you read it, what you do with it, and how you tell the story of your teaching to the people deciding your promotion. Sort before you feel. Close the loops. And when you write your annual self-evaluation, remember that a single rough semester is part of your evaluation data. You know better than most how to read a data set: one section doesn’t make the story, the pattern across the whole set does. Your teaching record deserves the same read.