Scholarly Publishing Curriculum for PhD Programs

Many students learn publishing by crisis, but a six session module can teach journal choice, authorship, peer review and open access on purpose.

Opinion · 10 min read · Directive Publications Editorial Team

Minimum curriculum (6 sessions)

1. How journals and peer review work
2. Choosing a journal / avoiding predatory outlets
3. Authorship, ORCID, affiliations
4. Writing for review (figures, claims, limitations)
5. Responding to reviewers
6. Open access, versions, and corrections

Session one is a map of the pipeline, walked once from end to end: submission, editorial checks, desk decision, assignment to a handling editor, reviewer invitations, reports, decision, revision rounds, acceptance, proofs, publication. Teach the vocabulary explicitly, because students hear it used and infer the wrong meanings — desk rejection, aims and scope, handling editor as distinct from publisher, single-anonymous and double-anonymous and open review, the difference between a decision letter and the reviewer reports it encloses. Spend real time on the decision categories. A student who reads "major revision" as a polite rejection withdraws a paper that was going to be published, and that misreading is common enough to be worth stating flatly: major revision means the editor wants the paper back. Say out loud what a normal timeline looks like in your field, and that silence during it is not a signal, so nobody emails the editor in week three asking whether something has gone wrong.

Session two has the largest downside if it goes untaught. Teach shortlisting as a mechanical procedure rather than a matter of taste: build the candidate list from your own reference list, since the journals you cite most are the journals whose readers need the result; read the aims and scope against your own abstract and cut anything that needs a paragraph of argument to make it fit; look at three recent papers in each candidate and ask whether yours would look at home beside them. Then teach verification. Check indexing in the databases your library subscribes to, not on the journal's own homepage, because an indexing claim is the cheapest thing on earth to print. Look up two or three editorial board members and see whether their own pages mention the role. Find the article processing charge, in writing, before submission. Check whether the journal has ever published a correction, because a journal that has never corrected anything either publishes nothing or fixes nothing. Name the recurring red flags plainly: a flattering solicitation email, a promised review turnaround measured in days, a scope that spans unrelated disciplines, and a fee that only becomes visible after acceptance.

Session three should happen before students have anything to argue about. Work through the widely used criteria — a substantial contribution to the work, involvement in drafting or revising it, approval of the final version, and accountability for it — then apply them to two or three cases from your own field rather than leaving them abstract. Practical mechanics matter as much as principle. Every student registers an ORCID in the session and connects it to their existing outputs. Everyone picks one form of their name and commits to it, because a record split across three name variants is tedious to merge and invisible until it costs someone a shortlisting. Affiliation is where the work was done, with a current address added if they have moved since. If the target journals use a contributor role taxonomy such as CRediT, use its role names from the start, so the submission form becomes a transcription rather than a negotiation. Close by naming the two situations students cannot handle alone — a senior figure expecting a slot they did not earn, and a colleague who leaves mid-project — and tell them who in the department they take those to.

Session four reframes writing around the two people who decide. The editor skims a title, an abstract and the figures before deciding whether to send the paper out at all; the reviewer reads the methods with one question in mind, which is whether they could tell what was actually done. So: one claim per figure, legends that can be read without the body text, axes labelled with units, and a title that states the finding rather than the topic. Then teach claims and limitations as a matching problem. The claim has to be the size of the design — an observational study yields association, a single site yields a result about that site — and the fastest way to teach it is to hand out a paragraph that overreaches and have the room rewrite it down to what the data support. A limitations section that names a limitation and says what it prevents is stronger than one that lists three and dissolves each with a reassurance. The most effective exercise here is to give students a published paper and have them write the review, since reading like a reviewer once permanently changes how a person writes.

Session five is a workshop, not a lecture. Give the room a decision letter with its reviewer reports and have them number every comment as the reviewers laid it out, splitting any paragraph that bundles three requests into three items. Sort into easy, hard and disagree. Then teach the response format: quote the comment, answer it, and point to the section and line numbers in the revised file where the change lives. Every disagreement rests on evidence or a stated reason, never on the claim that something is standard in the field. Give them two rules to keep. Volume is not difficulty — a letter with thirty comments can be a week of work and a letter with four can be two months. And the editor's letter outranks the reviewers: if the editor says the statistics must be addressed and the rest is at your discretion, that is the priority order regardless of how forcefully a reviewer wrote. Cover the mundane parts too, since they cause real delays: how to ask for a deadline extension before it expires, and what to do when two reviewers ask for opposite changes, which is to do one, say why, and let the editor arbitrate.

Session six is about the record after acceptance, and it is the session most often skipped. Students should leave able to name which version of their paper they are holding — the preprint, the accepted manuscript, or the published version of record — because the answer determines what they may post, where, and when. Cover licence choices in plain terms, funder and institutional deposit requirements, embargo periods, and where to look for a waiver or discount on a publication charge. Then cover corrections, which almost nobody teaches. A correction is the normal, expected mechanism for fixing an honest error in the published record, and it is distinct from a retraction. A researcher who knows that is far more likely to report a mistake they found in their own paper than to hope nobody notices. Say plainly that authors initiate corrections, that editors expect them, and that the reputational risk runs the opposite way from what students assume.

Core links for reading lists:

• Predatory journal checklist
• Authorship
• Peer review process
• Directive Academy

Who should teach it

Pair someone who publishes actively in the students' field with your scholarly communication librarian. The researcher supplies the judgement calls and the real decision letters; the librarian supplies indexing, licences, repositories and deposit rules, and usually already runs a version of session six for the whole institution. Use anonymised material from your own department wherever you can — a real cover letter, a real review, a real solicitation email out of somebody's spam folder — because generic examples do not stick, and students assume the bad cases happen elsewhere.

If you genuinely cannot find six slots, run sessions two and three. Journal choice and authorship are where the irreversible damage happens: a paper published in an outlet the field does not recognise cannot be moved, and an authorship dispute discovered at proof stage can cost a year. The other four sessions cover mistakes that are expensive but recoverable.

Assessment that works

• Critique a fake predatory homepage
• Draft a cover letter
• Write a sample response to a hostile review comment

All three are practical, because the failure mode of publishing training is a student who can recite the criteria and still submits to the wrong place. Keep them short, about an hour each, and mark them against what a student would actually do at submission.

For the homepage critique, give an archived page rather than a live link, and ask for a written verdict naming three specific signals plus the checks they would run to confirm each one. Mark the checking, not the verdict. A student who says the journal looks suspicious because the design is poor has learned nothing; a student who says the indexing claim needs verifying at the database and the board members need looking up has learned the transferable skill. Include an ambiguous case if you can — a small, legitimate, unglamorous journal — so the exercise is not just pattern matching on ugliness.

For the cover letter, set the constraints hard: one page, addressed to a named journal, one sentence on what is new, one on why it fits this journal in particular, the corresponding author named with an institutional email that will still work in two years, and the required declarations. Mark it with a single test — could an editor desk-reject this letter for telling them nothing? The most common failure is a reworded abstract, and it is worth showing the room what that looks like before they write their own.

For the reviewer response, hand out one genuinely sharp comment and require a three-part answer: what you understood the reviewer to be asking, what you changed or why you are not changing it, and where the change lives in the revised manuscript. Ask for the venting draft as well, written first and discarded on purpose. Writing the letter you cannot send and then rewriting it teaches the difference far better than being told to be professional.

What to leave out

The module fails when it expands. Leave out metrics and rankings beyond the one sentence that a journal's usefulness is whether the right readers see the work. Leave out the compliance lecture on misconduct definitions, which belongs in research integrity training and will crowd out everything practical. Leave out any tour of a specific submission system's interface, since it will have changed by the time they use it. And leave out grant writing, however tempting the overlap looks. The module covers the decisions a student makes between having a result and having a published paper, and nothing else.

Outcome

Judge the module by what students do at submission, not by how they rate the sessions. The signs that it worked are specific and observable: shortlists built from reference lists instead of inboxes, authorship conversations that happen when the analysis plan is fixed rather than the week of submission, revisions that go back inside the deadline, and students who can tell you which version of their own paper they just posted. The sign that it did not work is a first submission that arrives on a supervisor's desk as an emergency.

Graduates submit with fewer avoidable delays — and fewer integrity surprises. Explore For Authors.