Preparing a submission

What does a journal actually require before it will consider a paper?

More than the manuscript. Take PLOS ONE, one journal that publishes its requirements in full. It requires submissions in English and reporting guideline checklists for several study types. It also requires the data underlying the findings to be deposited in an appropriate public repository, unless already supplied with the article.¹

Its data requirement is specific rather than general. Data should be at least as open as a Creative Commons attribution license and deposited before or on publication.¹

Every journal publishes its own version of that list. The one that matters is the one belonging to the journal you are submitting to, and reading it in full before formatting anything saves the most time.

Desk rejection often turns on an item from that list rather than on the science. A paper returned unread has usually failed a condition of consideration rather than a judgment of quality.

What is a reporting guideline, and do I need one?

A reporting guideline is a checklist setting out what a paper of a given type has to report. They exist because readers and reviewers cannot assess what is not described, and they are now required rather than recommended by many journals.

PLOS ONE names several by study type. Clinical trials follow CONSORT with a completed checklist supplied, and systematic reviews and meta-analyses supply a completed PRISMA checklist and flow diagram. Observational studies conform to STROBE, Mendelian randomization studies use STROBE-MR, and animal research follows ARRIVE.¹

The number of guidelines in existence is larger than most authors realize. The EQUATOR Network, which maintains the library of them, lists over seven hundred, searchable by study type, clinical area, and section of the report.² Finding the right one is a research task rather than a formatting task.

Do I have to format the paper to the journal before submitting?

Read the instructions rather than assuming either way. Some journals now accept format-free initial submissions and ask for formatting only on acceptance, and others reject at the desk for it.

What is almost never optional is the material that goes alongside the manuscript. Checklists, data statements, competing interest declarations, ethics approvals, and trial registration numbers are conditions of consideration rather than formatting preferences.

An editor can format a paper to a journal's specification where that specification is supplied. What an editor cannot do is guess which of several journals you have in mind.

Reformatting for a second journal after a rejection is a second job, and it is worth budgeting for. Authors who expect one round of submission are the ones most surprised by the cost of the third.

What goes in a cover letter?

What the paper is, why it belongs in this journal, and anything the editor has to be told. That last category includes related submissions, prior presentation of the work, and any conflict a reader should know about.

The letter is not a summary of the abstract. An editor reading it is deciding whether to send the paper out, so the useful content is what makes it a fit for this journal's readers rather than a restatement of the findings.

Length is short by convention and by kindness. A cover letter running past a page is competing for attention with the paper it introduces.

One further thing belongs in it where the journal asks: suggested or excluded reviewers. Journals differ on whether they want these, and the instructions say which.

What about data availability and ethics statements?

They are checked, and a missing one stops the submission. PLOS ONE requires data underlying reported findings to be in a public repository unless already provided with the article, and requires deposit before or on publication.¹

Statements of this kind have a further property worth knowing: they are read against the paper. A data availability statement promising deposit at a repository that does not hold the data is a problem discovered later rather than avoided.

An editor checks that these statements exist, are internally consistent, and say what the journal's template asks them to say. Whether the data are actually where the statement says they are is not something an editor can verify.

The same holds for ethics approvals and trial registration numbers. An editor confirms that a number is present and formatted correctly, and cannot confirm that it belongs to this study.

What happens if I submit to a different journal next?

Everything has to be checked again, because the requirements differ. Reference style, structure, word limits, the abstract format, and the required declarations all change between journals, and a paper carrying the previous journal's conventions announces itself.

The traces authors miss are usually in the apparatus rather than the prose: a reference list in the old style, an abstract with the old headings, and a cover letter naming the previous journal.

This is a genuinely good use of a short editorial pass. Conforming a manuscript to a new specification is mechanical work, and it benefits from somebody who has not read the paper forty times.

It is also work with a definite end, which makes it easy to quote. A conformance pass is priced differently from an edit and is worth asking for by name.

Do the checklists have to be filled in?

Where a journal asks for a completed checklist, yes, and completed means completed. PLOS ONE asks for the checklist supplied as supporting information for the study types it names.¹

The item that catches people is the location column. PRISMA's own 2020 checklist carries a column headed for where each item is reported, so a checklist submitted without page or section references has not been completed in the sense intended.

An editor can check that every item is addressed somewhere in the paper and that the locations given point at the right places. Whether the underlying reporting is adequate is a reviewer's judgment rather than an editorial one.

Doing that check before submission also improves the paper. Items a checklist asks for and the manuscript does not contain are found while there is still time to add them.

Who checks the references?

The authors, with help. Reference lists are where late errors concentrate, because they are assembled from several sources over a long period and get edited last of all.

Copy editing includes checking cross-references and building the consistency record that a reference list is checked against.³ An editor can confirm that every in-text citation has an entry, that every entry is cited, that the style is uniform, and that dates, volumes, and author spellings are internally consistent.

What an editor cannot confirm is that a reference says what the sentence claims it says. That check belongs to the authors, and it is the one that matters most. Every document a client sends us is treated in confidence, including unsubmitted manuscripts and data statements.