
Before editing: freeze the study and collect the journal rules
Begin after the study design, analysis, author-approved conclusions, and paper structure are stable. Editing too early creates avoidable rework because polished sentences may be deleted when the argument changes. Save a dated master copy, confirm who can approve changes, and work on a duplicate.
Download the current instructions for authors from the target journal. Build a short requirements sheet covering article type, word limits, abstract format, heading levels, reference style, figure specifications, supplementary files, anonymisation, and required declarations. The journal's own instructions take priority over a generic style guide.
If the paper still needs stronger section logic, clearer claims, or extensive sentence revision, begin with research manuscript editing. If content and language are settled, a final academic proofread may be the more appropriate stage.
Create one source of truth
Keep the manuscript, author details, journal checklist, figures, supplementary files, and response notes in one version-controlled folder. Rename the approved file only after the final PDF check.
Research paper editing checklist by section
Edit each section for the job it must perform. A sentence can be grammatically correct and still be in the wrong section, overstate the evidence, or repeat information already visible in a table. The checks below apply broadly, but the journal's article structure remains decisive.
| Section | Editing questions | Common final check |
|---|---|---|
| Title | Does it identify the topic and study clearly without unsupported claims? | Match the title in the manuscript, submission form, cover letter, and running head. |
| Abstract | Does it accurately represent the purpose, methods, principal results, and conclusion? | Follow the required structured or unstructured format and word limit. |
| Introduction | Is the problem, relevant context, gap, and study purpose easy to follow? | Make the final paragraph's objective consistent with the abstract and methods. |
| Methods | Could an informed reader understand what was done from the supplied description? | Use consistent names for groups, measures, software, dates, and approvals. |
| Results | Are findings reported without interpretation that belongs in the discussion? | Cross-check values, units, sample labels, tables, figures, and supplementary material. |
| Discussion | Does interpretation follow from the reported evidence and acknowledge limitations? | Avoid repeating the results section or introducing unreported findings. |
| Conclusion | Does it answer the stated purpose at the right level of certainty? | Remove new evidence and claims broader than the study supports. |
Edit the argument, paragraph flow, and research language
Read once for logic before correcting individual sentences. Mark the function of each paragraph in a few words. Adjacent paragraphs should have distinct purposes, and the first sentence should orient the reader without merely repeating a heading. Move or combine material only when the change preserves the author's intended argument.
Then edit at sentence level. Prefer precise subjects and verbs, remove unnecessary repetition, and keep terminology stable. Check tense by function rather than applying one tense mechanically: methods describe completed actions, results report findings, and established knowledge may use the present tense where appropriate.
- Define abbreviations at first use in the main text and again where a table or figure must stand alone.
- Use the same term for the same concept, participant group, variable, instrument, and outcome throughout.
- Check that words such as causes, proves, improves, and prevents do not exceed what the design supports.
- Remove vague pointers such as this, it, and these when the intended noun is unclear.
- Separate author queries from direct corrections so factual decisions remain visible.
Cross-check every table, figure, and in-text callout
Tables and figures are part of the argument, not decoration. Read each item independently, then trace every in-text callout. Number items in the order first mentioned, confirm that the cited item exists, and check that the narrative does not conflict with the displayed values.
Captions and legends should explain the item without forcing the reader to search the main text for basic definitions. Verify units, decimal places, symbols, group labels, statistical notation, colour meaning, and footnotes. Remove duplicated data only when the journal discourages showing the same result in both a table and a figure.
For a deeper long-document audit, use the tables, figures, and references checklist.
Match in-text citations and the reference list in both directions
Run two checks: every in-text citation should have a reference-list entry, and every reference-list entry should be cited in the manuscript unless the journal permits a separate bibliography. Compare author names, years, letter suffixes, and group-author names exactly.
Apply the target style consistently to authors, title capitalization, journal or book details, volume, issue, pages or article numbers, DOIs, and URLs. A language edit can identify visible inconsistencies, but authors should verify that each source exists, supports the associated statement, and is represented accurately.
Do not invent missing details
Flag uncertain references for author verification. Never guess a DOI, page range, author, publication year, ethics identifier, or source title to make an entry look complete.
Check ethics, funding, authorship, conflicts, and data statements
Use the journal's required headings and wording for ethics approval, consent, funding, competing interests, author contributions, data availability, acknowledgements, and use of third-party material. Confirm each statement with the responsible author; an editor should not infer disclosures from the manuscript.
The International Committee of Medical Journal Editors maintains recommendations for authorship, disclosure, and reporting, while journals and disciplines may impose different requirements. Treat external guidance as a prompt for author verification, not as permission to manufacture missing statements.
Audit the complete journal submission package
The manuscript is only one part of the submission. Compare the final title, abstract, keywords, author order, affiliations, corresponding-author details, funding information, and declarations across the manuscript and submission portal. Follow the journal's rules for anonymous review; remove identifying details only where required.
- Main manuscript in the required editable or PDF format
- Separate title page, if requested
- Figures at the specified dimensions, resolution, colour mode, and file type
- Supplementary data, appendices, reporting checklists, and permissions
- Cover letter and highlights or graphical abstract, only when requested
- A clean file with comments, hidden text, and tracked changes handled according to journal instructions
Finish with separate editing, proofreading, and file checks
Do not ask one hurried read-through to solve every problem. Complete a logic pass, a language pass, a consistency pass, and a final proofread. After accepting changes, update automated fields, contents lists, references, and cross-references, then export the exact file type the journal requires.
Read the exported file page by page. Check line breaks, missing symbols, cropped figures, changed fonts, blank pages, hyperlink behaviour, and the order of end matter. Finally, compare the submission portal fields with the approved manuscript and keep a copy of everything submitted.
If you want a human review before upload, submit the research paper securely or compare editing and proofreading before choosing a service.
- 01Freeze content and record the target journal's current rules.
- 02Edit section purpose, argument flow, and claims.
- 03Edit sentences, terminology, tone, and consistency.
- 04Cross-check values, tables, figures, citations, and references.
- 05Verify declarations and every submission file with the responsible author.
- 06Proofread the approved manuscript and inspect the exported file.
Questions about this topic
Should I edit or proofread a research paper before submission?
Choose editing when section logic, argument flow, claims, or sentence clarity still need work. Choose proofreading when content is approved and the remaining task is a final language, consistency, and presentation check. Some papers need both stages in sequence.
Can an editor guarantee journal acceptance?
No. Editing can improve clarity, consistency, and presentation, but acceptance depends on research quality, journal fit, editorial screening, peer review, ethics, and decisions outside an editor's control.
Who should verify references and research facts?
The authors remain responsible for source accuracy, data, factual claims, ethical disclosures, and the final submitted version. An editor can flag inconsistencies and missing information without inventing or confirming facts they cannot verify.
