4.3 Submission Preparation and Reviewer Response¶
What you will learn on this page
- What reviewers focus on and how to prepare
- How to improve your title and Abstract
- How to pre-empt common reviewer comments
- How to respond to reviewer comments (Response to Reviewers)
- How to write a cover letter
- AI-use disclosure templates (English and Japanese)
- How to keep a work log
- Practical criteria for ethical AI use
The reviewer’s perspective¶
Reviewers evaluate papers under time constraints. They often judge quality using the following points.
- Abstract clarity: can the whole study be grasped quickly?
- Introduction logic: is the flow from gap to purpose coherent?
- Methods reproducibility: is there enough information to replicate?
- Objectivity of Results: are results and interpretation separated?
- Depth of Discussion: comparison with prior work and awareness of limits
- English quality: frequent errors can undermine trust
What reviewers decide in the first 10 minutes¶
| What they read | Time | What they judge |
|---|---|---|
| Title | 10 seconds | fit and interest |
| Abstract | 1–2 minutes | significance, novelty, concreteness of results |
| Quick scan of figures/tables | 2–3 minutes | quality of evidence and presentation |
| Last paragraph of Introduction | 1 minute | clarity of RQs and strength of gap |
| End of Conclusion or Discussion | 1 minute | what can and cannot be claimed, future direction |
A pre-submission 'first 10 minutes' check
Before submission, read only: Title → Abstract → Figures/Tables → last Introduction paragraph → Conclusion. Then ask AI to evaluate the first impression.
Prompt: simulate reviewer first impression
You are a reviewer for an international journal.
Read the title, Abstract, and research questions below and evaluate:
(1) Novelty (high/medium/low)
(2) Can the overall study be understood from the Abstract? (yes/partly/no)
(3) Are the RQs clear and testable? (yes/partly/no)
(4) Would you want to keep reading, and why?
Title: [Title]
Abstract: [Abstract]
Research questions: [RQs]
Improving the title and Abstract¶
Title¶
The title is your paper’s signboard and strongly affects discoverability.
| Type | Example | Note |
|---|---|---|
| Informative / declarative | AI-Assisted Writing Improves L2 Academic Vocabulary Use | hints at results, often attracts attention |
| Indicative | Effects of AI-Assisted Writing on L2 Academic Vocabulary | states topic only |
Prompt: propose improved titles
Informative (declarative) titles may increase citations
A study reported that papers with short informative titles that describe results are cited more often (at least in some fields).
Paiva, C. E., da Silveira Nogueira Lima, J. P., & Paiva, B. S. R. (2012). Articles with short titles describing the results are cited more often. Clinics, 67(5), 509–513. https://pmc.ncbi.nlm.nih.gov/articles/PMC3351256/
The title of Mizumoto and Teng (2024) is "Large language models fall short in classifying learners' open-ended responses," and one of the reviewers commented:
Perhaps the authors could consider rephrasing the title
to reflect the study's contributions and implications.
For example, something along the following lines:
Evaluating the limitations of large language models
in classifying learners' open-ended responses
for self-regulated learning.
In response, we cited the study above to argue that a declarative title may attract more citations, and persuaded the reviewer to keep the original form (whether it has actually led to more citations remains to be seen…).
Abstract¶
Please evaluate the Abstract below from a reviewer’s perspective.
(1) Is the purpose clear?
(2) Does it include essential method information?
(3) Are the main results concrete (including numbers where appropriate)?
(4) Does the conclusion follow from the results?
For items that need improvement, provide a one-sentence revision policy.
[Abstract]
For move structure and conventional phrases, see:
2.2 Checking Conventional Expressions with Corpora
Common reviewer comments and how to pre-empt them¶
By anticipating frequent reviewer comments and addressing them before submission, you can increase your chances.
| # | Common comment | Section | Pre-emptive action |
|---|---|---|---|
| 1 | “The novelty/contribution is unclear.” | Introduction | state the gap and contribution explicitly |
| 2 | “The literature review is insufficient.” | Introduction | cover key studies and synthesize |
| 3 | “The methodology lacks detail for replication.” | Methods | run the reproducibility checklist |
| 4 | “How were participants selected?” | Methods | state inclusion/exclusion criteria |
| 5 | “Effect sizes should be reported.” | Results | add effect sizes consistently |
| 6 | “Results and Discussion are mixed.” | Results/Discussion | separate reporting and interpretation |
| 7 | “The discussion is superficial.” | Discussion | compare with prior work and deepen implications |
| 8 | “Limitations are not addressed.” | Discussion | add a limitations subsection |
| 9 | “The English needs improvement.” | Whole paper | run grammar checks and editing |
| 10 | “References are outdated.” | Whole paper | confirm enough recent work is included |
Prompt: pre-review as a strict reviewer
How good is AI for 'review-like' feedback?
A study reported a fairly high correlation (r = .80) between LLM feedback and human peer review feedback (Liang et al., 2023). In a comparison of GPT-4-generated review comments with human ones (n = 308), 82.4% of respondents rated the AI feedback as “more helpful than feedback from some human reviewers.”
https://arxiv.org/abs/2310.01783
However, do not upload confidential unpublished manuscripts without careful consideration. Many publishers prohibit using AI to perform peer review, so use AI for your own writing support, not for official reviewing.
Responding to reviewer comments¶
If your paper receives a Revise and Resubmit (R&R) decision, you must prepare a Response to Reviewers.
Core principles¶
- Respond to every comment (missing one can be fatal)
- Start with appreciation (even if the comment is harsh)
- Specify exactly what you changed (page and line numbers if possible)
- If you disagree, explain your rationale politely
Example response format¶
Reviewer 1, Comment 1:
"The novelty of this study is not clearly stated."
Response:
We thank the reviewer for this valuable comment. We have revised
the Introduction (p. 3, lines 15–22) to explicitly state the
novelty of the present study. Specifically, we added a paragraph
clarifying that, while previous studies focused on X, the present
study is the first to examine Y in the context of Z.
Response templates by situation¶
| Situation | Template |
|---|---|
| You agree and revised | “Thank you for this suggestion. We revised the manuscript as follows: … Please see p. X, lines Y–Z.” |
| You partly agree | “We appreciate this comment. While we agree that …, we believe that … because …. However, we added … to address the concern.” |
| You disagree | “We thank the reviewer for raising this point. We respectfully disagree because …. We added a clarification on p. X ….” |
Prompt: classify and organize reviewer comments
Read the reviewer report below and classify each comment.
Categories:
(A) content revision needed (design, analysis, interpretation)
(B) writing revision needed (style, grammar, presentation)
(C) additional analysis/data needed
(D) author should rebut (reviewer misunderstanding or disagreement)
For each comment:
- category label
- one-sentence summary
- difficulty (high/medium/low)
- priority
[Reviewer report]
Prompt: draft a response paragraph in English
Draft a response paragraph in English for the reviewer comment below.
Requirements:
- confident but polite tone
- order: appreciation → response → location of change
- if revised, leave page/line as [TO FILL]
- if not revised, provide rationale
Reviewer comment:
[Comment]
Response policy (in Japanese):
[agree/partly agree/disagree + reason]
Using AI to draft responses
Responding to reviewer comments requires substantive research judgment. Use AI for English drafting support, but decide what to accept and what to rebut yourself.
Keep emotions out
Avoid phrasing like “The reviewer is wrong.” Use polite disagreement such as “We respectfully disagree.” Drafting with AI can help remove emotional wording.
Cover letter¶
Many journals request a cover letter. It is a document for the editor that explains your manuscript and submission intent.
What to include¶
| Element | Content | Note |
|---|---|---|
| Addressee | Editor-in-Chief name (check journal site) | “Dear Editor,” is acceptable if unknown |
| Manuscript title | the exact title | match the manuscript |
| Brief summary | what the study did (2–3 sentences) | a compressed Abstract |
| Contribution | what is new and why it matters | focus on value for journal readers |
| Fit | why this journal | relate to scope |
| Declarations | no dual submission, author agreement | follow venue rules |
| Statements | conflicts, AI disclosure if required | follow policy |
| Contact | corresponding author info | — |
Prompt: draft a cover letter
Using the information below, draft a cover letter for journal submission in English.
Requirements:
- formal academic tone
- within 300 words
- emphasize contribution and fit
- avoid excessive humility
- include an AI-use disclosure line if needed
Target journal: [journal]
Editor name: [name] (if unknown: "Dear Editor-in-Chief")
Manuscript title: [title]
Study summary (Japanese): [2–3 sentences]
Main findings (Japanese): [1–2 sentences]
Why this journal (Japanese): [1–2 sentences]
Authors: [names]
AI-use disclosure: [what you used AI for, if any]
Differentiate your cover letter
A cover letter is a chance to help the editor decide “this is worth sending to review.” State why the paper fits this journal specifically.
A personal rule: let AI draft only the summary part
After finishing a manuscript, rewriting the whole summary again can be exhausting. Cover letters are mainly for editors, so letting AI draft only the summary part is a practical personal rule.
AI-use disclosure and work logs¶
The biggest risk in AI-assisted writing is not the English quality, but being unable to explain later what you did yourself and what AI supported. Fix responsibility on the author side and keep a verifiable record.
Decide one rule sentence first
Example: I use generative AI only for structure consultation, paraphrase options, grammar explanations, alternative wording, and checking. I do not use it to add content, generate facts or citations, or produce sources.
Disclosure templates (English)¶
Place disclosure near Acknowledgements or in the cover letter. Replace the bold parts.
A. Language editing only (conservative):
The author used a generative AI tool (ChatGPT, OpenAI) to improve clarity and readability of the manuscript, including proofreading and wording suggestions. All content, analyses, interpretations, and final decisions remain the author's responsibility.
B. Including outlining and paraphrase options (practical):
The author used a generative AI tool (ChatGPT, OpenAI) for language editing and drafting support, such as outlining, paraphrasing options, and grammar checking. The author verified all statements and references and is fully responsible for the final manuscript.
C. Including Methods wording support:
The author used a generative AI tool (ChatGPT, OpenAI) to assist in converting technical procedures into natural-language descriptions and to refine wording. The author ensured accuracy of all methodological descriptions and verified all factual claims and citations.
D. Including code generation support:
The author used a generative AI tool (Claude, Anthropic) for drafting support (language editing, paraphrasing, outlining) and for generating initial analysis code in R/Python. All generated code was reviewed, tested, and verified by the author. The author is fully responsible for the accuracy of all analyses, interpretations, and the final manuscript.
E. Multiple tools used:
The author used the following AI tools during manuscript preparation: ChatGPT (OpenAI) for language editing and paraphrasing, Claude (Anthropic) for structural feedback, and Grammarly for grammar checking. All content, analyses, and interpretations are the author's own. The author verified all factual claims and references and takes full responsibility for the final manuscript.
How to choose a template
Work logs¶
You only need a level of detail that lets you explain “when, what, and how much.”
| Date | Section | AI use | Input type | How output was used | Your verification |
|---|---|---|---|---|---|
| 2026-02-08 | Introduction | paraphrase options | your English (5 sentences) | selected one option and edited | checked meaning drift |
| 2026-02-09 | Methods | clarify procedures | bullet notes | adopted wording, kept numbers | checked against original data/code |
| 2026-02-10 | References | DOI search hints | reference list | used hints only | verified on Crossref |
Where to store logs
One page in a research notebook (Notion, Obsidian, Google Docs, etc.) is enough. It becomes very helpful for reviewer responses and institutional explanations.
A practical unit of logging
A good unit is “section × revision round.” You do not need to log every sentence.
Example change-log format¶
For cases where detailed logs are needed (thesis work, competitive grants):
## Change log: Introduction (Paragraph 2)
### 2026-02-11 Revision 1: logic
- Tool: Claude
- Prompt summary: check gap strength
- AI feedback: gap is too abstract; missing what and why
- My action: added a concrete description of the missing population/data
- Adopted? yes (judged feedback valid)
### 2026-02-11 Revision 2: style
- Tool: ChatGPT
- Prompt summary: shorten subjects; adjust hedging
- AI output: 3 options
- My action: chose option 2 (active voice clarifies stance)
- Adopted? partial (combined with my wording)
Practical criteria for ethical AI use¶
The ethical boundary is not always simple. The spectrum below helps you position your own use.
An AI-use spectrum¶
| Level | Use | Typical evaluation |
|---|---|---|
| 1 | spell check, proofreading | ✅ generally fine |
| 2 | paraphrase options, style adjustment | ✅ often acceptable |
| 3 | outline consultation, structure checks | ✅ disclosure recommended |
| 4 | translation (Japanese → English) | ⚠️ disclosure needed |
| 5 | drafting sections with AI | ⚠️ disclosure required; debated |
| 6 | generating interpretation/discussion content | ❌ often not allowed |
| 7 | generating data/results | ❌ research misconduct |
Three tests when unsure¶
- Transparency test: Can you disclose this use honestly to the venue? If not, avoid it.
- Substitution test: If a human translator/editor did the same task, would it be acceptable? If yes, it is often acceptable.
- Responsibility test: Can you verify the output yourself? If not, do not use it.
Five principles for ethical AI use¶
| # | Principle | Practical action |
|---|---|---|
| 1 | Content is yours, expression is supported | keep claims, analysis, interpretation under your control |
| 2 | Do not let AI create sources | verify via databases |
| 3 | Final decisions are yours | do not accept outputs uncritically |
| 4 | Disclose and log | record and disclose scope of AI use |
| 5 | Do not reduce learning opportunities | use AI to support improvement, not replace it |
For detailed policies and examples, see:
1.3 Guidelines from Publishers and Educational Institutions