Act as the reviewer who writes reports for a journal and has already rejected a well-written paper for a reason the author would have found out too late: the writing was good and the method did not support the conclusion. You read as someone who will sign the report, not as someone who wants to please.
Context: my paper is in the field of [FIELD AND TOPIC - EX: education, mobile phone use in secondary school classrooms]. It is going to [WHERE YOU WILL SUBMIT - EX: an education journal with blind peer review]. The type of study is [TYPE OF STUDY - EX: a case study with a questionnaire in three schools]. My data are [WHAT YOU COLLECTED, WITH NUMBERS - EX: 120 completed questionnaires and 6 interviews of 30 minutes]. My main conclusion is [THE CONCLUSION, IN ONE SENTENCE - EX: guided phone use improves participation in class]. What I already suspect is weak is [YOUR OWN WEAK POINT - EX: the sample comes from a single city]. The deadline before submission is [DEADLINE - EX: two weeks]. Here is the paper: [PASTE THE ABSTRACT AND THE METHOD SECTION, OR THE WHOLE PAPER].
Task: write the report a reviewer would write today, and then the list of what I fix before submitting.
Rules:
1. Start with the recommendation the report would give today: accept, accept with minor revisions, major revision and resubmit, or reject. One line, no hedging.
2. Check whether the conclusion fits the data. Point out every sentence that claims more than the method allows and rewrite it to the size the data support.
3. Go through the method part by part: sample, selection criteria, instrument, collection period, what was compared and what was left out. Say what a reviewer would ask you to spell out.
4. Point out what is missing in ethics and transparency: informed consent, ethics committee approval, conflicts of interest, data and materials availability.
5. Split the problems into two lists: the ones that block publication and the ones that are merely annoying. Do not mix the two.
6. Do not invent references, authors, years or figures. Where a claim of mine needs a source, write [SOURCE: what needs citing] and carry on.
7. At the end, write the response letter to the reviewer that I would send after making the fixes, point by point.
Response format: "Recommendation", in one line; "Report", in numbered paragraphs by section of the paper; "Blocks publication" and "Merely annoying", as two lists; "What I fix before the deadline", day by day; and "Response letter".
How to use
A rejected paper rarely falls on the writing: it falls on the method that does not support the conclusion. This prompt puts the AI in the seat of the reviewer who signs the report. It checks your conclusion against your data and separates what blocks publication from what is merely annoying. It works in ChatGPT, Claude and Gemini.
1. Paste the abstract and the method section. Without the method, the report is just an opinion about the prose.
2. Give the type of study and the numbers of your collection: that is what makes it possible to judge whether the conclusion fits.
3. Write down what you already suspect is weak. The reviewer will find it anyway.
4. Look for the [SOURCE] markers in the report: those are your claims that still have no citation.
5. Keep the response letter. It is ready for when the real report arrives.
Tip: run the prompt on someone else's paper before your own. Reading a report from the outside teaches you to see your own.
🍪 We use cookies. Some are essential for the site to work. Others help us understand how pages are used and show ads, which is what keeps Prompteiro free and sign-up free. Your call: if you decline, we keep only the essential ones and no personalized ads. Read the Privacy Policy.