Writing a review: be specific about the problem and specific about the fix

The simplest standard for a review report: write the review you would want to receive.

Specific, evidenced, and pointing towards a fix — not a global judgement the author cannot act on.

Structure of a report

• A short summary of the paper, showing you read and understood it.
• An overall assessment: the contribution, and the largest problem.
• Major comments, numbered.
• Minor comments, in a separate section.

Each comment needs

• A location: page, line, section.
• What the problem is and why it is a problem.
• A suggested direction, where you have one.

Separate two kinds of issue

• Issues where the conclusion does not hold — these must be fixed.
• Issues where you would have done it differently — mark these as suggestions.
• Mixing them leaves the author and editor unable to prioritise.

Tone

• Review the paper, not the author.
• Anonymity is not a licence for harshness.
• Rigour about content and respect in tone are independent choices.

When to decline

• You lack expertise in the methods used.
• You have a conflict of interest.
• You do not have time to read it properly — accepting and skimming is worse than declining.

One thing worth remembering

For every problem you raise, suggest a direction.

That is the difference between a review that improves the paper and a review that only demonstrates the reviewer is clever — and authors who receive the first kind remember it for a long time.

Câu hỏi thường gặp