Manuscript Submission Checklist for First-Time Authors

Submitting your first paper is less about perfection than about not missing the basics, so work through this checklist before you click Submit.

Checklist · 8 min read · Directive Publications Editorial Team

1. Confirm the paper is submission-ready

1. Research question and contribution are clear in the abstract
2. Methods are reproducible enough for a peer to understand
3. Figures/tables are final, labeled, and cited in text
4. References are complete and consistent
5. Word count and article type match the journal

The abstract test. Give the abstract alone to a colleague in an adjacent field and ask them to tell you, in one sentence, what you found. If they cannot, an editor scanning the submission will not either. The contribution should be a claim, not a topic: not "we investigated X in Y," but "X does Z under condition Y."

Reproducible enough. The standard is whether someone with your skills, and no access to you, could repeat the study. That means software and version, instrument settings, inclusion and exclusion criteria, how missing data were handled, which statistical test you used and why, and where the data or code sits.

Figures and tables. Every figure must be cited in the text, in order, and every panel label in the caption must exist in the image. Captions should stand alone — a reader who sees only the figure should understand it. Check that axis labels and units are legible at the size the journal will print them, not at the size on your screen.

References. Every in-text citation appears in the list, and every entry in the list is cited. Reference manager output drifts over time, so look for duplicate entries, missing page ranges, and preprints cited as though they were journal articles. Check that nothing central to your argument has been retracted.

Length and type. Journals count words differently — some exclude the abstract, references and captions, some do not. If you are well over the limit, that is a structural problem, not a trimming problem.

The mistake to avoid is submitting a draft you know is unfinished in order to "get feedback." Peer review is not a writing service. An unready paper usually comes back as a rejection, and you have spent months to learn what a co-author could have told you in an afternoon.

2. Choose the right journal first

Match scope, audience, and article type — not only reputation talk. Verify peer review, fees, editorial board, and ethics policies.

Choose before you do the final formatting, because the journal decides your word limit, section structure, reference style, and whether you need a separate blinded file. Formatting first and choosing second means doing the work twice.

A fit test that works. Find three papers published in that journal in the last couple of years that you would genuinely cite in your own introduction. If you cannot find them, your paper will read as off-topic to the handling editor, however good it is.

What to verify before you commit. An editorial board of named people whose affiliations you can trace. A stated peer review model. Fees stated plainly on the site before submission, not disclosed after acceptance. Indexing claims checked at the index itself rather than on the journal's own page. A real contact address, and registered DOIs on published articles.

Guide: How to Identify a Predatory Journal · Browse: Directive Publications journals

3. Lock authorship before upload

1. Author list and order agreed in writing
2. Corresponding author named
3. Contributions recorded (e.g. CRediT)
4. All co-authors approve the final file

The submission system asks for the author list on the first screen, and that list becomes the record. Changing it afterwards usually requires written confirmation from everyone involved, including anyone being removed, and editors reasonably treat late changes as a signal that something was not agreed.

Do it in one email. Send the proposed list, the proposed order, who will be corresponding author, a one-line description of what each person did, and a date by which to object. Ask for a reply from every person. Silence is not agreement, and a co-author who never answered is exactly the co-author who raises a problem after acceptance.

Record contributions properly. A taxonomy such as CRediT gives you a fixed list of roles — conceptualisation, methodology, analysis, writing, and so on — that you assign per person, with more than one person allowed per role. Doing this while everyone remembers who did what settles arguments before they start, and most journals now ask for it anyway.

Name the corresponding author carefully. That person handles correspondence, proofs, and questions after publication, so use an institutional email address that will still work after the current contract ends. The role is about availability, not seniority.

People who helped but do not meet authorship criteria — funding acquisition alone, general supervision, a technician who ran a standard assay — belong in the acknowledgements, and you should ask their permission before naming them there.

Read: Who Counts as an Author?

4. Prepare the submission package

1. Title page / blinded manuscript (if required)
2. Cover letter
3. Highlights or graphical abstract (if required)
4. Data availability / ethics / funding / conflicts statements
5. Suggested reviewers (independent, institutional emails)

Blinding is more than deleting the names. If the journal runs double-anonymous review, the blinded file must also lose affiliations, acknowledgements, funding details, and self-references written in the first person ("in our earlier work"). Cite your own previous papers in the third person. Then check the file properties, because word processors store the author name in the document metadata whether or not it appears on the page.

The statements are not boilerplate. Data availability should name the repository and the identifier, or state plainly why the data cannot be shared and what a reader can request instead. Ethics approval should give the committee and the approval reference, plus a consent statement where people were involved. Funding should use the grant numbers in the form your funder requires. Conflicts should include the ones you think are too minor to matter — an undisclosed interest found later is a far bigger problem than one stated up front.

Suggested reviewers. Independent means no recent co-authors, no one at your institution, no current grant collaborators, and no one you thanked in the acknowledgements. Journals usually state their own window for what counts as recent, so check. Give institutional email addresses only; suggested reviewers with free webmail addresses are a known pattern in peer review fraud, and editorial offices check them.

Reviewer guidance: Suggesting Reviewers

5. Final technical check

1. Similarity screening reviewed thoughtfully
2. File names and formats match journal instructions
3. ORCID and affiliations correct

Read the matched sources, not the score. Overlap concentrated in methods boilerplate, standard instrument descriptions and the reference list is ordinary. The genuine problem is a run of your own sentences traced to a single source, or a passage lifted from your own earlier paper without quotation or citation. Fix it by rewriting from your notes rather than by shuffling words, and quote and cite where the original wording matters.

Technical checks are where submissions bounce. Line numbers, spacing, figure resolution and file format, figures embedded versus supplied separately, tables as editable text rather than images: journals check these before an editor reads anything, and a miss costs you a round trip.

ORCID and affiliations. Connect your ORCID through the submission system rather than typing the number, so the record links properly. List the affiliation where the work was done, and add a current address separately if you have since moved.

Before you upload, put every file in one folder named for the journal, in the order the system asks for them. Then submit, and check for the confirmation email and manuscript number — a submission that never reached the system is a submission that is not under review.

What to do if you have already clicked Submit

If you spot an error before the paper is assigned for review, email the editorial office once, quote the manuscript number, say plainly what is wrong, and ask whether you may upload a corrected file. Most offices will let you.

Do not fix it by submitting a corrected version as a new submission — that reads as duplicate submission and is treated seriously. If you need to move to another journal, withdraw formally and wait for confirmation before submitting anywhere else. A paper under consideration at two journals at once is a genuine ethics problem, not a paperwork slip.

See: How Similarity Screening Works · Peer Review Process

Bottom line

Work through the five stages in order, because each one constrains the next: the journal decides the format, the format decides the files, and the author list decides who has to approve them. Doing them out of order is what turns a week of preparation into a month.

A clean first submission saves weeks. When ready: For Authors · Submit Manuscript · more guides at Directive Academy.