A feedback comment is the sentence or two a teacher or lecturer leaves on one answer. It takes time to write and is not sure to be used: in one study, fewer than half the students collected theirs.
This guide covers the comment itself: what it is for, which kinds students act on, a three-part pattern, a comment bank and the whole-class comment.
EdProcto, which publishes this guide, prints your comment under its question on the student's result. The last section says where it is typed.
What is a written comment for, beside the mark and the right answer?
A comment tells the student how to get from the answer they gave to the answer that was wanted. The mark says how far the answer fell short. The right answer says where it should have arrived, which is enough for a fact. The comment is for a method, a definition or a paragraph, where the right answer does not show a student their own gap.
"Some good points, more detail needed" repeats the mark in words.
Every source was read on 10 October 2026, most as an abstract, which we mark. A study that another of our guides uses is cited here for a different finding.
Do comments work better without a grade beside them?
Often, though leaving the mark off does not make the comments good. In the classic experiment, R. Butler (1988), pupils given comments alone did best, as our timing guide sets out. Its abstract adds an exception: high achievers given grades kept their interest while more grades were expected (abstract read).
Smith and Gorard (2005) followed a comprehensive school that tried it. Of 104 Year 7 pupils in four mixed-ability groups, one had fuller comments and no marks for a year; the others had marks with minimal comments. The comment-only group's progress was "substantially inferior" (abstract read).
What should a comment be about: the task, the process or the person?
The task and how it was done, never the person.
- Kluger and DeNisi (1996) pooled 607 effect sizes from 23,663 observations. Feedback improved performance on average (d = 0.41), yet over a third of the interventions lowered it, and it helped less the more it drew attention to the self and away from the task (abstract read).
- Hattie and Timperley (2007) reviewed the evidence and proposed a model (abstract read). As Wisniewski et al. (2020, read in full) set it out, feedback addresses the task, the process, self-regulation or the self, and the last, usually praise of the person, is the least effective.
What praise does
Praise that names something in the work is information; praise of the student is a verdict.
Lipnevich and Smith (2009) crossed detailed feedback with a grade and with a statement of praise on an essay sat by 464 college students. Praise softened the harm a grade did, and detailed feedback still worked best with neither (abstract read).
How many comments, and in what form?
Few, each ending in an instruction.
- Ackerman and Gross (2010) report that marketing students took a large amount of feedback on an assignment badly, no better than no comments at all. They advise a moderate number of clear, specific comments (abstract read; it gives no sample size).
- Sugita (2006) wrote statements, imperatives and questions between the drafts of students of English as a foreign language. Across 115 changes, imperatives moved students to revise more than the other two (abstract read). One small study; still, follow a question with an instruction.
Spoken or written?
Neither has been shown to teach more. Voelkel and Mello (2014) gave half an undergraduate class audio feedback on a low-stakes essay and half written. Audio took five minutes longer a script and left students more satisfied; neither form raised marks in the later exam essay (abstract read). In the meta-analysis by Wisniewski et al. (2020), the channel (oral, written, or video, audio and computer) made no significant difference.
Do students read their comments, and what makes them respond?
Many do not, and asking for a response is the most promising lever we found.
- Sinclair and Cleland (2007) had 360 third-year medical students at Aberdeen sign for the formative feedback on their essays: 46% collected it, higher achievers more often (abstract read).
- Jackson and Marks (2016) had master's students write assessed reflections on their feedback and, in a second year, withheld grades as well. 77% favoured the reflections and 57% the withholding, which made some anxious. Class grades rose over both years; the abstract reports no comparison group. Our timing guide has a randomised trial of feedback shown before the grade.
How do you write a comment in three parts?
Say what the answer did, what is missing and what to do next, in that order, in about two sentences. The third part carries it: feedback on where to go next is the kind most received and interpreted (Wisniewski et al., 2020).
| Work | What was done | What is missing | What to do next |
|---|---|---|---|
| A calculation: 15% of 240, answered 3,600 | You multiplied the right two numbers. | The division: per cent means out of 100. | Estimate first: 10% of 240 is 24. Then redo it. |
| A definition: opportunity cost, answered "the cost of something you buy" | You tied the term to a choice about spending. | The alternative given up, which is the whole idea. | Rewrite it using "given up", with an example from your week. |
| A paragraph on a poem | Your point is clear and you quote the line. | No sentence says how the quotation supports the point. | Add a sentence after the quotation, beginning "This shows". |
Four rules:
- Few comments (Ackerman and Gross, 2010): one an answer, on the gap that costs the most marks.
- Name the answer, never the student (Kluger and DeNisi, 1996).
- End on an instruction (Sugita, 2006), one that takes five minutes.
- Praise only what you can point to.
Is a bank of reusable comments worth keeping?
Yes, for the gaps that recur, with the first part still written from the answer in front of you.
Denton et al. (2008) asked 169 first-year pharmacy students about feedback on a laboratory report; 110 had been marked by tutors using handwritten annotation or software that assembles a feedback report. The software's reports were rated higher on four features, among them what was deficient and what was done well, and the tutors reported taking less time (abstract read).
The risk is the comment that fits everyone. Adalberon (2021) found most of 411 examiners' comments in a Norwegian teacher-education programme formulaic and close to the grade descriptors (abstract read). So bank the second and third parts only, filed by the error:
- "The unit is missing. Write it beside every answer in this section."
- "This describes where the question said explain. Add the reason, starting with because."
How long should a comment take, and when does one whole-class comment beat thirty?
About a minute an answer is our working rule, untested, and a whole-class comment wins once the same sentence has gone to a third of the class. No study we read timed written comments on short answers.
Thirty answers at a minute each is half an hour on one question. If ten share a gap, it was not yet taught. Write it once:
Most answers stopped at lactic acid. The extra oxygen you breathe in afterwards is what breaks it down.
Put it in the question's explanation, or open the next lesson with it and have each student rewrite their own answer, as in our guide to the questions they got wrong. We found no experiment setting a whole-class comment against single ones: ERIC returned four records for "whole-class feedback" on 10 October 2026, none that comparison.
Which comment fits which kind of work?
Match the comment to what the right answer cannot show.
| Kind of work | The comment | In EdProcto |
|---|---|---|
| Multiple choice, one word, a number | None one by one: an explanation written with the question | "Explanation shown to the student afterwards" |
| A calculation with working | One line at the first wrong step | "Note for the student" |
| A short answer or a definition | Three parts, two sentences | A note, yours or a first draft from Procto AI |
| A paragraph or longer | One strength, the largest gap, one next step | A note; leave the slips |
| One gap across the class | A whole-class comment | The explanation, or the next lesson |
For the rubric behind a comment, see short answer questions and the single-point rubric.
A worked example: five answers to one question
The question: "Explain why your breathing rate stays high for some minutes after a sprint. (3 marks)" One mark each for: the muscles respired anaerobically, without enough oxygen; lactic acid built up; extra oxygen is needed afterwards to break it down. The answers are invented.
| Answer | Marks | Comment |
|---|---|---|
| "Because you are tired and need to get your breath back." | 0 | You say the body must recover, which restates the question. Name how the muscles released energy without enough oxygen. |
| "The muscles did not get enough oxygen so they respired anaerobically. You breathe fast to get the oxygen back." | 1 | Anaerobic respiration, with its cause: one mark. Missing: what it produces. Add the product, then finish "the extra oxygen is used to". |
| "Lactic acid builds up and makes the muscles ache. Breathing fast gets rid of it by breathing it out." | 1 | Lactic acid builds up: one mark. It is not breathed out. Rewrite your second sentence to say what oxygen does to it. |
| "Your heart beats faster to pump more blood and glucose to the muscles." | 0 | True, and it explains heart rate. The question asks about breathing. Answer again, starting from what the muscles ran short of. |
| "The muscles needed energy faster than oxygen arrived, so they respired anaerobically and made lactic acid. Afterwards extra oxygen breaks it down, so breathing stays fast." | 3 | All three steps, in order. Next: say why a trained runner's breathing settles sooner. |
Four of the five never say what the extra oxygen does. That is the whole-class comment above, and with it each single comment shrinks to a line.
Where does a comment go in EdProcto, and what does the student see?
EdProcto is the best tool for comments that get used, because your note is printed under its question on the student's result, with the right answer and the explanation, and you decide when that result is released. In Sinclair and Cleland (2007), fewer than half the students collected feedback that had to be fetched.
- Where it is typed. On the grading page, every answer waiting for a reader has "Points" and "Note for the student", then "Save my grade". The note is one line of text, and the server keeps up to 4,000 characters.
- On an answer the rules marked. "Add a note, or set the points by hand" opens the same boxes. Leave the points, type the note and press "Save my note". The mark stays the rules' own, so "Re-mark this quiz" marks it again from a corrected key and keeps your note.
- A first draft. On a typed open answer of a shape it grades, "Ask Procto AI" returns the points and a note of up to 2,000 characters: one to three sentences on what was right and what was missing, or, where your rubric is set out in lines, a short line for each criterion with its points before them. The next step is yours to add. It is saved under Procto AI's name for you to keep or change: edit it, press "Save my grade", and the mark is yours. See checking the marks it proposes.
- What is kept. Each mark is stored with its note and who made it: the rules, Procto AI or you.
What the student sees
The result lists the quiz "Question by question" while "Show a question-by-question breakdown" is on. Under a marked question come "Answer:" with the right answer (for an open question, your model answer), "Explanation:", then your note as plain text, with the points at the side or "to be graded" while an answer waits. The list prints the question and not what the student wrote, so open the note with their own words.
The "Feedback" group of the settings decides when.
- "Show the correct answer" and "Show the explanation" each take "At once", "After submitting" or "Never".
- "Show the score" adds "When released". Until you release, the score, the marks, your notes, the right answers and the explanations stay on the server.
- A note travels with the marks. With the score at "Never", the comment that reaches the student is the explanation.
A student opens the result again from the quiz's link, on the device they handed in on, with "See your result"; a note added since is there.
The comment written once, for everyone
Each question in the builder has "Explanation shown to the student afterwards", with the grey prompt "Why the answer is right, in plain words." After marking, add the class's commonest gap to it: a result is built afresh each time it is opened, so it carries the new wording from then on.
Notes by hand and these settings are on every plan. A quiz takes thirty students on Free, 100 on Tutor and 500 on Educator, which adds the full analysis; Procto AI's grading comes with the Tutor plan. See pricing and the page for teachers.
Quick questions
How long should a feedback comment be?
About two sentences on one gap. Ackerman and Gross (2010) found that a large amount of feedback was received no better than none.
Should every answer get a written comment?
No. Answers marked by rule get one explanation written with the question, and open answers a note each: in EdProcto, "Explanation shown to the student afterwards" and "Note for the student".
Is praise in feedback useful?
When it names something in the work. In Lipnevich and Smith (2009), detailed feedback worked best with no praise and no grade beside it.
Is audio feedback better than written feedback?
Students often like it more, and attainment has not followed: Voelkel and Mello (2014) found no gain in a later exam essay.
References
- Ackerman, D. S., & Gross, B. L. (2010). Instructor feedback: How much do students really want? Journal of Marketing Education, 32(2), 172–181. https://doi.org/10.1177/0273475309360159
- Adalberon, E. (2021). Providing assessment feedback to pre-service teachers: A study of examiners' comments. Assessment & Evaluation in Higher Education, 46(4), 601–614. https://doi.org/10.1080/02602938.2020.1795081
- Butler, R. (1988). Enhancing and undermining intrinsic motivation: The effects of task-involving and ego-involving evaluation on interest and performance. British Journal of Educational Psychology, 58(1), 1–14. https://doi.org/10.1111/j.2044-8279.1988.tb00874.x
- Denton, P., Madden, J., Roberts, M., & Rowe, P. (2008). Students' response to traditional and computer-assisted formative feedback: A comparative case study. British Journal of Educational Technology, 39(3), 486–500. https://doi.org/10.1111/j.1467-8535.2007.00745.x
- Hattie, J., & Timperley, H. (2007). The power of feedback. Review of Educational Research, 77(1), 81–112. https://doi.org/10.3102/003465430298487
- Jackson, M., & Marks, L. (2016). Improving the effectiveness of feedback by use of assessed reflections and withholding of grades. Assessment & Evaluation in Higher Education, 41(4), 532–547. https://doi.org/10.1080/02602938.2015.1030588
- Kluger, A. N., & DeNisi, A. (1996). The effects of feedback interventions on performance: A historical review, a meta-analysis, and a preliminary feedback intervention theory. Psychological Bulletin, 119(2), 254–284. https://doi.org/10.1037/0033-2909.119.2.254
- Lipnevich, A. A., & Smith, J. K. (2009). Effects of differential feedback on students' examination performance. Journal of Experimental Psychology: Applied, 15(4), 319–333. https://doi.org/10.1037/a0017841
- Sinclair, H. K., & Cleland, J. A. (2007). Undergraduate medical students: Who seeks formative feedback? Medical Education, 41(6), 580–582. https://doi.org/10.1111/j.1365-2923.2007.02768.x
- Smith, E., & Gorard, S. (2005). "They don't give us our marks": The role of formative feedback in student progress. Assessment in Education: Principles, Policy & Practice, 12(1), 21–38. https://doi.org/10.1080/0969594042000333896
- Sugita, Y. (2006). The impact of teachers' comment types on students' revision. ELT Journal, 60(1), 34–41. https://doi.org/10.1093/elt/cci079
- Voelkel, S., & Mello, L. V. (2014). Audio feedback – better feedback? Bioscience Education, 22(1), 16–30. https://doi.org/10.11120/beej.2014.00022
- Wisniewski, B., Zierer, K., & Hattie, J. (2020). The power of feedback revisited: A meta-analysis of educational feedback research. Frontiers in Psychology, 10, Article 3087. https://doi.org/10.3389/fpsyg.2019.03087