Turning a Thesis Into a Publishable Paper

A thesis proves to examiners that you can do research. A journal article tells busy readers one thing you found. Getting from one to the other is a rewrite, not a trim.

Guide · 10 min read · Directive Publications Editorial Team

What changes between a thesis and a paper

A thesis chapter and a journal article can report the same data and still be different documents. The chapter was written for examiners who had to read all of it and needed evidence that you understand the field and can justify every decision. The article is written for specialists who will read the abstract, look at one table, and decide whether the rest is worth their time.

So do not start by deleting paragraphs until the chapter fits a word limit. Trimming keeps the chapter's shape, and the shape is what is wrong. Start from what the paper needs to say, then take material from the chapter to build it.

Step 1: Find the one claim the paper makes

Before you open the chapter, write one sentence stating what the paper found. For quantitative work a workable template is: "In [population], [exposure or intervention] was associated with [outcome] compared with [comparator]." For qualitative work, name the setting, the participants and the central finding.

Then test the chapter against that sentence:

• Every result you keep should support, qualify or limit the claim.
• Every paragraph of the introduction should explain why the question matters or what was not known.
• If you need "and" to join two unrelated findings, you probably have two papers, or one paper plus supplementary analyses.

If you cannot write the sentence, the chapter may not hold a standalone finding yet. Combine it with another chapter or finish the analysis it needs, rather than submitting and hoping reviewers find the point.

How many papers the thesis holds

Count questions, not chapters. The general introduction and general discussion exist to connect the chapters, so they seldom stand alone as research papers. Each empirical chapter is a candidate if it answers a distinct question with a complete result.

The line to watch is splitting. The ICMJE Recommendations say manuscripts based on the same data set should add substantially to each other to warrant separate publication, and should cite the earlier papers from that data set. A simple test: if two planned papers share participants and methods, and a reader would need both to understand either one, write one paper. Where two manuscripts are genuinely separate but closely related, ICMJE asks authors to say so in the submission letter and supply copies of the related material.

Step 2: Choose the journal before you rewrite

The target journal decides the article type, the length, the abstract format, the reference style and whether it considers thesis material at all. Rewriting before you know those means rewriting twice. If you have not settled on a journal, start with how to choose the right journal.

For your first choice, write down:

• the word limit for your article type, and how words are counted (ICMJE describes a count of the main text that excludes the abstract, acknowledgments, tables, figure legends and references)
• limits on tables, figures and references
• the abstract format, which ICMJE notes differs from journal to journal
• the reporting guideline expected for your study design
• what the author guidelines say about theses, preprints and prior publication

Step 3: Rebuild the chapter instead of trimming it

Open a blank document with the journal's headings. For original research these are usually Introduction, Methods, Results and Discussion, the structure ICMJE calls IMRAD. Fill each section from the chapter with only what the claim needs, and write the abstract last, from scratch, in the journal's format.

Introduction

This is where the biggest cut happens. A thesis literature review shows breadth; a journal introduction gives context, identifies the gap and states the objective or hypothesis, and ICMJE advises citing only directly pertinent references. Remove the history of the field, definitions specialists do not need, frameworks the paper does not test, and anything that previews another chapter.

Methods

Keep what a reader needs to judge or repeat the study: design, setting, participants, interventions or exposures, outcomes and how they were measured, sample size reasoning, analysis and ethics approval. Cite a registered or published protocol rather than re-describing it. Move full search strategies, instrument development and extended sensitivity analyses to supplementary files where the journal allows them. ICMJE says supplementary material should be submitted with the manuscript so that it is peer reviewed alongside it.

Results

Keep the results that answer the question, including those that do not support your hypothesis, and give the main findings first. Watch for one common conversion error: ICMJE asks for data on all primary and secondary outcomes identified in the Methods. If you drop an outcome from the Results, drop it from the Methods too, and check that this is consistent with any registration or protocol.

Discussion

Open with the principal finding in a sentence or two, compare it with the most relevant evidence, state the limitations, and say what follows for research or practice. Delete the thesis-wide research agenda.

Then sweep the draft for leftovers: "as discussed in Chapter 2", first-person singular where co-authors now exist, abbreviations defined chapters earlier, and figures sized for a bound volume rather than a journal page.

Step 4: Settle authorship before a draft circulates

A thesis carries one name; the paper will usually carry several. The ICMJE authorship criteria require all four of these: substantial contributions to conception or design, or to acquisition, analysis or interpretation of data; drafting or critically revising the work; final approval of the version to be published; and agreement to be accountable for all aspects of the work. Everyone who meets all four should be an author, and ICMJE lists general supervision of a research group, on its own, as not qualifying.

That test cuts both ways. A supervisor who helped design the study, interpret the results and revise the manuscript, and who approves and stands behind it, qualifies in the ordinary way. A supervisor whose only role was oversight does not, however senior. Apply the same test to the statistician, clinical colleagues who recruited participants, and any co-supervisor.

Agree these points in writing. An email thread that everyone replies to is enough:

1. Who is an author, and the contribution that qualifies each person.
2. Author order, which ICMJE says the author group decides, not the editors.
3. The corresponding author, with an email address that will still work after you graduate.
4. Who drafts, who revises, and the deadline for final approval.

If agreement breaks down, ICMJE says journal editors do not arbitrate authorship conflicts; the institution where the work was done should be asked to investigate. Our guide to handling authorship disputes covers how to stop it getting that far.

Step 5: Add what a medical journal will ask for

Your thesis may never have needed some of the items an editorial office checks first. Check each one against the new manuscript, not the old chapter:

• Reporting guideline. ICMJE names CONSORT for randomised trials, STROBE for observational studies, PRISMA for systematic reviews and meta-analyses, and STARD for diagnostic accuracy studies. The EQUATOR Network lists others, including COREQ and SRQR for qualitative research. Fill in the checklist with the manuscript's own page numbers.
• Ethics approval. Name the approving committee and give the approval reference. ICMJE says authors should be prepared to provide documentation when editors ask.
• Consent. If the paper includes identifying patient details, ICMJE calls for written informed consent for publication, which requires the patient to be shown the manuscript.
• Trial registration. ICMJE recommends that journals require registration at or before the first participant's consent for enrolment, and states that retrospective registration, for example at submission, meets none of the purposes of registration. If your thesis includes a trial, confirm where and when it was registered before you choose a journal, and put the number at the end of the abstract, where ICMJE recommends journals publish it.
• Data sharing. ICMJE requires manuscripts submitted to ICMJE journals that report clinical trial results to include a data sharing statement. Other journals may ask for a data availability statement, so check the wording required.

Step 6: Disclose the thesis and handle the repository copy

The COPE position on publishing material from theses says a thesis generally consists of unpublished work, and that an author who owns the copyright may publish thesis material even if the thesis is freely available in a repository or carries a licence allowing reuse. Several large publishers take the same line. Elsevier's policies for authors, for example, list publication as an academic thesis among uses it does not view as prior publication, while noting that some society-owned titles and journals operating double-anonymous review have different policies.

Work through it in this order:

1. Read the target journal's own policy. If it treats an openly available thesis as prior publication, choose another journal or ask your graduate school whether the repository copy can be embargoed. Rules vary by institution.
2. Check copyright and co-workers. COPE's position assumes you own the copyright, and says anyone who co-authored part of the thesis must agree to the decision to submit.
3. Know which kind of thesis you have. COPE notes that papers already published and included in a thesis by publication cannot be republished as original work.
4. Cite the thesis. COPE says the thesis should be acknowledged as the source of the work and cited in the publication. For example: "This article is based on [author]'s doctoral thesis, [title], [university], [year], available at [repository link]." Place it where the journal's guidelines direct.
5. Tell the editor. Name the thesis and link to it in your cover letter. If the journal uses double-anonymous review, ask the editorial office how to cite it without identifying yourself to reviewers.

Disclosure also answers the similarity question before anyone asks it. Crossref Similarity Check compares manuscripts with published academic and general web content, so an open thesis can appear in the report as a match. A match to a thesis the editor already knows about is easy to account for, and the rebuild in Step 3 will have removed most verbatim overlap anyway. For how editors weigh those matches, see how to read a similarity report.

Before you submit

1. One-sentence claim written, and every section serves it
2. Journal chosen, with its word limit, abstract format and thesis policy checked
3. No cross-references to other chapters left in the text
4. Outcomes named in the Methods match the outcomes reported in the Results
5. Author list, order and corresponding author agreed in writing
6. Reporting checklist, ethics, consent, registration and data statements complete
7. Thesis cited in the manuscript and named in the cover letter
8. Related manuscripts from the same thesis disclosed to the editor

Start with the chapter that is closest to a single, complete answer, not the one you find most interesting. Once you have a target journal, read its author guidelines before you begin rebuilding. For Directive Publications journals, start at the For Authors page, and submit your manuscript when it is ready.