A teaching feedback form is one of the easiest things to send and one of the easiest to waste. The template exists, the link goes out, and three weeks later there is a spreadsheet with forty rows, most of them scoring four out of five, and two comments that contradict each other. Nothing in that spreadsheet tells anyone what to do on Monday.
The problem is almost never the question types. It is that the form was asked to do three unrelated jobs at once, that students had no reason to believe an honest answer would change anything, and that nobody was named as responsible for what came back. Those three things are fixable, and fixing them matters more than the wording of any individual question.
Decide which of three jobs the form is doing
There are three separate reasons a teaching feedback form gets sent, and they pull in opposite directions.
The first is to improve a course while it is still running. That form is short, asked early, about things the teacher can actually change, and its value expires in a fortnight. The second is to create a record: a score that goes into a departmental review, a comparison across terms, evidence for a promotion file or an accreditation visit. That form has to stay identical year on year, because the whole point is comparability. The third is to give a student a private channel for something specific: a reading list that costs too much, a lecture capture that has no captions, a seminar where one voice dominates. That form is not a survey at all. It is a request that needs a reply.
Most teaching feedback forms in circulation are trying to be all three. The result is a questionnaire that is too long to fill in honestly, too vague to act on quickly, and too inconsistent to compare with last year. It also fails the third job silently, because a student who reports a real problem in a free-text box at the end of a twenty-question survey has no way of knowing whether anyone read it.
Splitting them costs nothing. Two short forms and one permanent link are easier to build than one long form, and each can be answered in the state of mind it needs. The mid-course form can be blunt and disposable. The end-of-course form can be formal and unchanged. The permanent link, sometimes described as an open channel or a course contact form, sits in the syllabus all term and produces one message at a time. The use cases for that third kind look nothing like a survey, because each submission is a piece of correspondence rather than a data point.
Anonymity is a trade, not a virtue
Anonymity is usually treated as the setting that makes feedback honest. It does raise candour on questions students feel exposed answering, particularly anything about the teacher as a person rather than the course as a design. It also removes four things at once.
It removes the reply. A student who describes a genuine problem cannot be told what happened next, and the absence of any visible consequence is exactly what trains a cohort to stop answering.
It removes the count. One person writing the same complaint four times and four people writing it once look identical in an anonymous free-text column, and the two situations call for different responses.
It removes the follow-up question. Half of all useful feedback is unusable as written because a single detail is missing, and there is no way to ask which week, which reading, which room.
It removes the check. Whether the change made things better for the person who raised the problem is unknowable if that person cannot be identified.
There is a middle setting that is often better than either extreme: an optional contact field, with plain text next to it saying who will see the address and what it will be used for. Students who want a reply supply one. Students who want to be blunt leave it empty. The form stays honest and stops being a dead end.
One warning about small groups. In a seminar of nine, a comment that mentions a specific week or a specific piece of work identifies its author to anyone who was in the room, whatever the form says about anonymity. Promising anonymity that the group size cannot deliver is worse than not promising it, because the promise is what persuaded the student to write something identifying. Say instead that names are not collected and that comments may be recognisable in a small group.
And check the claim against the tool. Some forms record an email address by default, or require a sign in before a file can be attached, and a form that says anonymous while quietly logging an account has broken something that will not be repaired. Open the live link in a signed out browser, submit a test answer, and look at what actually arrives.
Ask about behaviour, not about liking
The single biggest gain in a teaching feedback form comes from replacing judgement questions with behaviour questions. Judgement questions return a number that cannot be acted on. Behaviour questions return a fact.
| Question as usually written | What comes back | Question that produces an action |
|---|---|---|
| Rate the quality of teaching, 1 to 5 | A 4, meaning nothing in particular | Which part of the last three sessions was hardest to follow, and what made it hard |
| Was the pace appropriate | Yes, from people with opposite views of the pace | Where did the pace stop working for you: the first half, the second half, or the problem sets |
| Were the materials useful | A 4 from everyone who downloaded nothing | Which handout or video did you actually go back to, and why that one |
| Any other comments | Silence, or one unrelated complaint | What is one thing to keep exactly as it is, and one thing to change before the next cohort |
Three practical rules follow from that table.
Use one scale and keep it
Mixing a five point agreement scale, a ten point satisfaction scale and a three point frequency scale in the same form makes the answers impossible to read side by side and makes the form feel longer than it is. Pick one scale, label every point in words rather than numbers alone, and use it for every scored item all term. Consistency matters more than which scale is chosen.
Keep the free text narrow and keep it to two boxes
An open box labelled comments collects whatever is closest to the surface, which is often the room temperature. Two narrow boxes collect what the course needs: one asking what to keep, one asking what to change first. Narrow framing produces shorter answers, and shorter answers are easier to act on and easier to read across a hundred submissions.
Cut the form to what will be read
Every item on a teaching feedback form should have a named person who will read it and a plausible decision it could change. Items that survive because they were on last year's form, and that nobody has looked at since, are the reason students believe the exercise is theatre. Cutting a twenty item form to eight usually raises both the response rate and the quality of the free text, because students who can finish in ninety seconds will write a real sentence at the end.
Send it when the answer can still change something
Timing decides what a teaching feedback form can be used for, and the three useful windows do different work.
A single session form, sent within a day of a specific class, gets detail that no end of term survey can recover. Students remember which example failed and which explanation landed, and they have not yet flattened the term into an overall impression. This is the right form for a new unit, a redesigned lecture, or a guest session.
A mid-course form, sent around the point where roughly a third of the teaching is done, is the only one that can change the experience of the people answering it. That is also what makes it worth answering. The cohort watches what happens next, and a visible change in week six buys honest answers for the rest of the term and for the term after.
An end of course form is a record rather than an intervention. Sent before results are released, it captures the course as experienced; sent after, it captures the course as remembered through a grade. Neither is wrong, but the two cannot be compared with each other, so pick one and keep to it.
The awkward truth about the end of term form is that it is the one most institutions require and the one students least believe in, because every change it produces lands on a cohort they will never meet. Telling them exactly that, and naming one thing the previous cohort changed, does more for the response rate than a reminder email.
Response rate is a design problem, not a motivation problem
Low response rates are usually blamed on apathy and are usually caused by friction. The fixes are dull and they work.
Give the form time in the session rather than sending it into a full inbox. Two minutes at the end of a class, with the link on screen, routinely beats a week of email reminders. Make sure it works on a phone, because that is what students have in their hands, and a form that shows one question per screen is easier to complete standing up than a long scrolling page. State the closing date on the form itself. Send one reminder, not four.
Then look at who did not answer. A thirty per cent response rate is not a small version of the cohort. It is the students at the two ends, the ones who were delighted and the ones who were angry, and the middle is missing. That skew is fine as long as it is named when the numbers are quoted. Reporting a mean score from thirty per cent of a class as though it described the class is how a teaching feedback form starts producing false confidence instead of information.
The part that is usually missing: what happens after it closes
Collecting answers is solved. Every tool does it well enough that the differences are cosmetic. What follows is where teaching feedback goes to die, and it has four parts.
Somebody has to own each item that needs action. When the same spreadsheet is shared by a course leader, a departmental administrator and two tutors, an unassigned comment is nobody's job by default. A form tool with response management gives each submission an owner and a status, which is the same information a shared spreadsheet holds in a column that everyone forgets to fill in.
Anything that needs a reply has to get one, and the reply has to be findable later. A message answered from a personal mail client leaves no trace next to the submission, so the next person to open that response cannot tell whether it was handled. Sending from the same screen as the response, and keeping the sent message attached to it, is the difference between a record and a recollection.
Decisions have to be recorded against the feedback that caused them. Two words next to an item, such as changed for next term or cannot change, room allocation, turn a survey into an audit trail. The following year it takes minutes rather than an afternoon to say what was done.
And the record has to survive the term. Feedback that lives in one person's spreadsheet leaves when they do. A tool that keeps the response, its owner, its status and the reply together, and exports all of that rather than the answers alone, is what makes a three year comparison possible. This is what separates a form builder from something with response management built in, and it is worth checking on a real submission rather than a feature list. A three question test form and one submitted response answer it faster than any brochure.
What to change first
Split the form by job: one blunt mid-course form, one unchanged end of course form, one permanent link for problems that need a reply. Then add an optional contact field to the first two so that an honest answer can be followed up instead of disappearing into a column. Once that is in place, check that each response can carry an owner, a status and the reply itself, with Halict or any tool built the same way.
Q1. Should a teaching feedback form be anonymous?
Anonymity raises candour on questions about the teacher personally, and it removes the ability to reply, to count how many people share a view, and to ask a follow up question. An optional contact field is often the better setting, since students who want a response supply an address and the rest leave it blank. In a group of nine or ten, be careful about promising anonymity that the group size cannot deliver.
Q2. How many questions should a teaching feedback form have?
Few enough that every item has a person who will read it and a decision it could change. Eight focused items with two narrow free-text boxes usually produce more usable material than twenty scored items, because students who can finish quickly still have the patience to write a real sentence at the end.
Q3. When is the best time to send a teaching feedback form?
Around the point where a third of the teaching is done, if the goal is to improve the course for the people answering. Within a day of the session, if the goal is detail about one specific class. At the end, if the goal is a record that can be compared with other terms, in which case send it at the same point relative to results every year so the comparison holds.
Q4. What can be done about a low response rate?
Treat it as friction rather than apathy. Give the form two minutes of class time with the link on screen, make sure it works on a phone, put the closing date on the form, and send one reminder rather than several. Then state the response rate whenever a score is quoted, because the students who answer first are usually the most and least satisfied rather than a small copy of the cohort.
Q5. What should happen to comments that cannot be acted on?
Record them as declined with a one line reason, such as timetabling or room allocation, rather than leaving them unmarked. An unmarked comment looks unread, which is the impression that stops the next cohort answering, and a short reason is also the fastest way to show a review panel that the feedback was considered.