HOW SINGAPORE EDUCATION WORKS · University peer-review responses
Find your next learning step
Copy every reviewer comment into a numbered matrix, state the action or reasoned disagreement, record exactly where the manuscript changed and run a final one-to-one completeness check.
Full chapter index · Practice and checking · Sources and next reading · How Singapore Education Works
Peer review becomes manageable when it stops being a long, emotional letter and becomes a traceable set of claims. A response matrix links each comment to a decision, a manuscript location and evidence that the revision was made.
Journal requirements control the real resubmission. This article is a general academic workflow, not a policy of every Singapore university or journal. Follow the editor’s letter, submission portal and field-specific conventions.
The choices at a glance
| What you notice | Likely process gap | Useful next move |
|---|---|---|
| Several comments are answered in one paragraph | Reviewer requests have been bundled | Give each distinct request its own row and identifier |
| The response says “done” | The revision is not locatable | Quote the new wording briefly and give page, section or line information |
| The author disagrees defensively | Claim and evidence are mixed with emotion | Restate the concern fairly, then give a concise evidence-based rationale |
Full chapter index
See the structure
Work and diagnose
Repair and practise
Recall, transfer and questions
Keep the editor’s decision letter and each reviewer report intact. Assign stable identifiers such as E1, R1.1 and R2.3. If one paragraph contains three requests, split it into three rows while retaining the original wording.
A useful matrix has at least four fields: comment, decision, change made, and manuscript location. Add an evidence or verification field when the revision involves analysis, data or a new source.
Do not paraphrase away the reviewer’s concern. The original language is the audit trail; your interpretation belongs in a separate column.
Most responses fall into three routes. Comply: make the requested change. Clarify: the manuscript already addresses the point, but the location or wording is inadequate. Disagree: retain the approach and provide a respectful, evidence-based reason.
| Decision route | What the manuscript needs | What the response needs |
|---|---|---|
| Comply | Visible revision | Exact location and concise description |
| Clarify | Stronger signposting or explanation | Show how the revision resolves the ambiguity |
| Disagree | Possibly no change, or a boundary statement | Fair restatement, evidence and reasoned rationale |
Assumptions: Reviewer 1 asks the authors to define the sampling frame, report missing-data handling and add a limitation about generalisability. The manuscript currently mentions the sample but not those details.
Create rows R1.1–R1.3. For R1.1, add the sampling-frame definition to Methods, paragraph 2. For R1.2, add the number of missing observations and the stated handling rule to Methods, paragraph 4. For R1.3, add a limitation explaining that findings may not generalise beyond the sampled institutions.
Checked response: all three comments have separate decisions, manuscript changes and locations. The authors reread the revised passages, verify that the missing-data statement matches the analysis record, and confirm that no response promises a change absent from the manuscript.
A response can quote a short new phrase, but the matrix is not a substitute for editing the manuscript itself.
If the matrix says “addressed”, ask whether a new reader can find the evidence. Replace vague completion language with a location and the substance of the change.
If a comment appears factually mistaken, first test whether the manuscript caused the misunderstanding. A clarification may be more useful than proving the reviewer wrong.
If new analysis is requested, record assumptions, software or code version, data exclusions and checked output. A response letter must not outrun the reproducibility record.
Use a five-part repair: quote the concern, name the decision, describe the manuscript change, identify the location and verify the revised claim against the source material.
For disagreement, avoid status language such as “the reviewer failed to understand”. Write: “We agree that the boundary needs to be clearer. We have retained X because Y, and added Z to Methods/Limitations.”
After repairing the row, check connected rows. One change may answer two comments, but each reviewer still deserves a direct response and cross-reference.
Guided triage
- Read the editor’s instructions before drafting responses.
- Split and number comments without deleting context.
- Assign an owner and a verification method to each row.
Independent audit
- Filter for rows without manuscript locations.
- Compare every promised change with the submitted revision.
- Check tables, figures, supplementary files and tracked changes where required.
- Ask a co-author to sample rows and reproduce the trace.
The final audit is one-to-one: every comment has a response, every promised revision exists, and every new claim is supported.
Before the next submission, rebuild the matrix columns from memory. Delayed recall turns a one-off administrative task into a reusable research habit.
Peer-review customs differ across journals, fields and countries. Some editors expect a point-by-point letter; some use portals or annotated files. The transferable core is respectful traceability, not one fixed document format.
Changed-context transfer: use the same matrix for an ethics amendment, supervisor feedback or a thesis examination report. The authority and rules change, but comment-decision-change-location remains useful.
Should authors accept every suggestion?
No. They should address every comment. A respectful disagreement may be appropriate when supported by evidence and consistent with the editor’s instructions.
Can several comments point to one change?
Yes, but respond to each comment separately and cross-reference the shared revision so no request disappears.
Should the response quote the whole revised paragraph?
Follow journal guidance. A short excerpt plus a precise location is often enough; long duplication can make checking harder.
Who should own the matrix?
One coordinator can maintain it, but every co-author responsible for a change should verify the relevant row. Authorship and approval rules still apply.
Sources and next reading
- NUS Libraries — What are peer-reviewed articles?
- NUS Lee Kong Chian Natural History Museum — Nature in Singapore instructions to authors
- NUS Medicine, The Asia Pacific Scholar — Author policies
Continue inside eduKateSG
- University reproducible research figures
- University data-cleaning decision logs
- University research file naming and version control