HOW SINGAPORE EDUCATION WORKS · Peer Review of University Writing
Find your next learning step
Choose the question closest to your situation, or read the full guide in order.
Full chapter index · Practice and checking · Sources and next reading · How Singapore Education Works
A useful peer review tells a writer what a reader understood, where that understanding broke down, and which revision deserves attention first. Read against the assignment brief, point to an exact passage, and explain the effect of the problem. The writer then decides how to revise and checks that the new version actually works.
This guide is for university students exchanging feedback on permitted drafts. It concerns classroom peer review, rather than journal refereeing or permission to collaborate on every assignment. Your module's rules determine whether, how and with whom a draft may be shared. The examples below are fictional; no private student work is reproduced.
The choices at a glance
| What you notice | Possible issue | Useful next move |
|---|---|---|
| The main claim is unclear | The reader needs an orientation | Paraphrase the claim before detailed comments |
| The calculation is correct but the conclusion is not | The inference needs checking | Compare measured evidence with the claim |
| Reviewers disagree | Different uncertainties may be involved | Separate requirements from preferences |
| Comments do not lead to change | The revision decision is missing | Record an action and a verification step |
Full chapter index
Define the review and read the reasoning
Check an example and handle disagreement
Revise, practise and check independently
Transfer across disciplines and answer questions
CHAPTER 1 OF 8 · Define the review and read the reasoning
1. Give the reader a defined job
Before exchanging drafts, establish the task, intended audience and stage of writing. A reader checking the structure of an early argument should not spend the whole session polishing punctuation. A final proofreading request has a different purpose.
Send the permitted brief or a short description of its criteria alongside the draft. Add one focused question: “Does paragraph three explain why the comparison supports my claim?” is easier to act on than “Is this good?” If the review itself is assessed, follow the module's instructions for format and submission.
Read the complete draft once before commenting in detail. Summarise its main claim in your own words. The gap between that summary and the writer's intended claim may identify the most valuable revision. The purpose is to test what the text communicates without relying on the writer's live explanation.
NUS CELC's discussion of peer reviewing treats review as an opportunity to examine communication and develop the reviewer's own judgement. That is a teaching perspective from NUS, not a rule that every Singapore university uses the same review process.
CHAPTER 2 OF 8 · Define the review and read the reasoning
2. Comment on the reasoning before the polish
Check the central relationship between claim and evidence. What is being asserted? What information supports it? How does the draft explain the connection? A citation can identify a source while the reasoning that connects it to the claim remains missing.
Then examine the sequence. Does the reader encounter the necessary definition before it is used? Does a paragraph contain one identifiable move, or does it begin with a comparison and end with an unrelated recommendation? Identify the point where you lost the thread, rather than declaring the whole draft confusing.
Make comments specific and bounded. “In paragraph four, I cannot tell whether the percentage describes all respondents or only those who completed the course” identifies an answerable uncertainty. “Bad statistics” does not.
Preserve strengths as well. Tell the writer which definition, example or transition helped you understand. This is practical information about what to keep during revision. A draft can become weaker if the writer removes a clear explanation while responding to a vague instruction to shorten everything.
Language corrections still matter. Deal with a repeated error or an ambiguity that changes meaning, and indicate its location. Avoid rewriting whole paragraphs in your own voice. The writer needs to practise making the decision and remain responsible for the submitted text.
CHAPTER 3 OF 8 · Check an example and handle disagreement
3. A worked case: a survey claim that goes too far
Assume a fictional draft describes an informal survey of 20 volunteers from one university club. Twelve say they prefer recorded briefings to live briefings. There is no random sampling, no comparison of actual learning outcomes, and no information about students outside that club.
The draft concludes: “Recorded briefings improve learning for all university students because 60% prefer them.” Check the arithmetic first: 12 divided by 20 is 0.60, or 60%. The calculation is correct. The conclusion remains unsupported.
A useful review comment is: “Your percentage correctly describes 12 of the 20 club volunteers. I cannot see evidence here about learning outcomes or all university students. Could you limit this sentence to the preference reported in this group, then explain what additional evidence a learning claim would need?”
One defensible revision is: “In this informal survey, 12 of 20 club volunteers preferred recorded briefings. This describes the respondents' stated preference; it does not establish better learning or represent the university population.” The writer can then discuss how a separate study might investigate learning, if the assignment calls for it.
The review succeeds by locating the inferential problem, not merely changing “all” to “many.” “Many students learn better” would still claim an outcome the survey did not measure. Check the revised sentence against the actual evidence, rather than assuming that a softer adjective fixes the logic.
CHAPTER 4 OF 8 · Check an example and handle disagreement
4. Handle disagreement and uncertainty honestly
A reviewer can misunderstand a passage, overlook a definition or make an incorrect suggestion. Treat a comment as evidence about a reading experience, not automatic authority over the subject. The writer should inspect the passage and the relevant sources before deciding.
When reviewers disagree, separate the issues. One may be asking for a clearer definition; another may be challenging the claim itself. Both can be addressed without treating the number of votes as a verdict. For a disciplinary convention or a disputed requirement, consult the brief, an approved reference or the instructor.
Distinguish verification from impression. “I could not follow this step” reports your experience. “The cited source does not contain this result” requires checking the source. If you have not opened it, say that the source claim still needs verification.
Respect the permitted sharing arrangement. Do not upload someone else's draft to a public service or AI tool merely because that seems convenient. Obtain the relevant permission and follow the institution's rules. For broader authorship and source questions, use the guide to academic integrity, sources and AI.
CHAPTER 5 OF 8 · Revise, practise and check independently
5. Turn the comments into a revision decision
After receiving a review, make a short revision list. Start with the issue that most affects the assignment's purpose. A missing comparison may deserve attention before several awkward phrases. Estimate a manageable next action, such as rewriting one claim and checking its supporting paragraph.
For each substantial comment, record a decision: revise, verify, or retain with a reason. “Retain” is not a refusal to learn if the evidence and task support the original choice. It should be an informed judgement rather than an automatic defence of the first draft.
In the survey case, the decision is to narrow the claim and separate preference from learning. The action is to rewrite the conclusion and check every earlier use of “improves.” The verification is to reread the paragraph as a new reader and confirm that it no longer generalises beyond the sample.
Save versions according to your normal course arrangements. A short note about what changed can support later reflection. It also makes a consultation more useful: “I narrowed the population, but I am uncertain how to discuss selection bias” gives the instructor a specific learning question.
This process complements using feedback and making a next attempt, with the extra responsibility of evaluating another student's reasoning fairly.
CHAPTER 6 OF 8 · Revise, practise and check independently
6. Practise reviewing, then check your own draft
For guided practice, review this invented sentence: “Eight of ten interviewees liked the workshop, so the workshop increased their skill.” Assume the only question asked was whether they liked it. Identify one correct part and one unsupported step.
Checking answer: eight out of ten is 80%, and the sentence can report the stated liking among those interviewees. Enjoyment is not a measurement of skill change. A repair should narrow the claim or introduce appropriate evidence of skill, rather than inventing results.
For an independent task, review: “Both libraries had 40 visitors on Tuesday; therefore they serve their communities equally well.” Name two uncertainties before proposing a revision. Possible checks include opening hours, population served, visitor purposes and what “serve well” means. Equal visitor totals alone cannot establish equal service quality.
Now inspect one paragraph of your own draft with the same questions. State the claim, locate the evidence and identify what bridges them. If the bridge exists only in your head, add the explanation. If the evidence does not support the bridge, narrow the claim or investigate further. Do not fill the gap with confident wording.
CHAPTER 7 OF 8 · Transfer across disciplines and answer questions
7. Carry the skill across borders and disciplines
MIT OpenCourseWare's account of feedback in a 2017 writing course describes guided peer comments that attend to structure, logic and places where readers lose understanding. This is one course example, not a requirement for all MIT or Singapore students.
The meaningful global connection is the shared problem of writing for someone who cannot read the author's mind. A reader from another discipline may need a term defined; a reader from another country may not recognise a local institution or acronym. That difference can reveal assumptions worth explaining.
Transfer the skill to a policy memo or a technical report. The audience and evidence standards change, but the reviewer still needs to locate the claim, inspect support and explain the effect of a gap. A scientific methods section and a literary interpretation do not use identical criteria. Use the discipline's actual conventions rather than forcing every text into the survey example.
Later, recall one valuable comment without opening the old review. Explain the reasoning behind it, then find a different passage where the same issue might occur. That checks whether the learning has become a reusable judgement rather than a correction tied to one sentence.
CHAPTER 8 OF 8 · Transfer across disciplines and answer questions
8. Common questions about peer review
What if I am not an expert?
You can accurately report what you understood and where the text became unclear. Mark subject questions that need verification. You do not need to pretend to know a source or a technical fact in order to provide a useful reading response.
Must I accept every suggestion?
Follow any formal review requirements, but evaluate substantive suggestions against the task and evidence. Explain your decision where a response is requested. The writer remains responsible for the final argument.
Should friends exchange complete answers?
Only within the assignment's permitted collaboration rules. Peer review is not general permission to share restricted work, supply a submission for copying or rewrite another student's assessment. Ask the module team when the boundary is unclear.
How do I avoid sounding harsh?
Locate the passage, describe the reading difficulty and propose a useful question or action. Discuss the text rather than the writer's ability. Specific feedback can be both candid and respectful.
Sources and next reading
Checked 22 October 2026. The NUS CELC peer-review discussion supplies Singapore teaching context. MIT OpenCourseWare: Providing Students with Feedback supplies the course-specific international comparison. The numerical cases and review comments are original fictional examples, not reported research findings.
Continue with academic reading, arguments and evidence or return to the education hub.