How to publish a scientific article

To understand how to publish a scientific article, it helps to see the entire process at once: select a journal, adapt the manuscript to its requirements, submit the files, pass editorial screening, respond to reviews, revise the text and, only after acceptance, go through production and online or issue publication. A mistake at the first step can cost more than a weak cover letter: a paper that does not fit a journal's scope may receive a rapid rejection regardless of the quality of the research.

Journal selection begins with scope

Do not look only at the journal title or its metrics. Read the aims and scope, then inspect the latest 10–20 papers. If the journal never publishes work using a similar method, research object or type of question, take that seriously. The relevant criteria are the readership, accepted article types, length limits, data policy and ethical requirements.

Check indexing in the primary databases rather than relying on a badge on the journal website. For open-access journals, verify the title in DOAJ when that is relevant. For Scopus or Web of Science, use the official source lists if indexing matters to your institution or funder. A journal appearing in a search engine is not the same as official indexing.

Manuscript structure

For original research, IMRaD is common: Introduction, Methods, Results and Discussion. ICMJE recommendations describe this structure as a reflection of the logic of a study, not as decorative formatting. The introduction defines the problem and objective. Methods explains how the research was conducted. Results reports what was found. Discussion interprets the findings and their limitations.

Not every article type should be forced into IMRaD. Reviews, case reports, theoretical papers and protocols may use other structures. The instructions of the journal you have chosen take precedence over a generic template.

What to check before submission

The title must match what the study actually demonstrates and should not imply causation when the design shows only an association. The abstract should accurately represent the data in the paper. Tables and figures should be numbered consistently and cited in the text. Authorship, conflicts of interest, funding, ethics approval and study registration must be declared where required.

If AI-assisted tools were used, read the journal's policy. Current ICMJE recommendations require authors to disclose the use of AI-assisted technologies where applicable and make clear that such systems cannot be listed as authors. Human authors remain responsible for the accuracy, integrity and originality of the work.

A cover letter without ritual filler

A cover letter should help the editor evaluate the submission. It normally identifies the manuscript, explains why it fits the journal, confirms that there is no prohibited duplicate submission and discloses material conflicts if the submission system does not collect them elsewhere. Depending on the journal, you may also need to identify related manuscripts or prior dissemination.

Avoid claims such as “our revolutionary study will transform the field.” Two concrete sentences about the research question, the data and why the journal's readership should care are stronger.

What happens after you click Submit

A submission may first undergo a technical check covering files, declarations and formatting. The editor then decides whether the manuscript should be sent for review. A desk rejection can result from a poor scope match, limited novelty, a fundamental design problem or failure to follow submission requirements.

If the paper goes to reviewers, possible decisions include acceptance, rejection, minor revision, major revision or revise and resubmit. Journals use different labels. Read the editor's letter separately from the individual reviewer reports, because the editorial decision governs the next step.

How to respond to reviewers

Prepare a response document with columns or clearly separated sections for the reviewer comment, your response, the change made and the location of that change. Answer every substantive point, including the comments you disagree with. A disagreement should be supported by data, method or journal policy, not by “we think otherwise.”

When a reviewer identifies a genuine weakness, correct it and say exactly what changed. If a requested change is impossible or inappropriate, explain why and, when necessary, add the issue as a limitation in the manuscript.

Timelines and APCs

The interval between submission and publication varies by months across journals and disciplines. Treat promises such as “publication within 48 hours” with caution when a journal claims to conduct substantive peer review. If the journal publishes decision-time statistics, use them. Reports from individual authors can be useful as secondary context, not as a guarantee.

An APC, article processing charge, may be zero, fixed or dependent on article type. Gold open access is often financed through APCs, but not always. Diamond open access normally means that neither readers nor authors pay for access or publication. The models are explained in more detail in article No. 22, “Open access: gold, green, diamond.”

Checklist before submission

• the journal's scope matches the topic and research design;
• structure and length requirements are met;
• the abstract contains no findings that are absent from the paper;
• tables and figures have appropriate rights, numbering and captions;
• authorship and conflicts of interest are declared;
• ethics approval and registration are stated where required;
• indexing claims have been checked in official databases;
• publication charges and the open-access model are understood before submission;
• the cover letter is specific and does not simply repeat the abstract;
• the manuscript is not simultaneously submitted where journal policy prohibits it.

Build a short journal list before the manuscript is finished

Choosing a journal on the final day forces you to reshape a finished paper around accidental requirements. A better approach is to identify three to five candidates once the study and article type are clear. For each candidate, note four things: scope fit, intended readership, accepted article type and restrictions on length, tables or supplementary materials.

Do not reduce the choice to a single metric. A high citation indicator does not compensate for a poor thematic fit. A rapid decision is not useful if the journal lacks the indexing required by your institution. Zero publication fees do not automatically mean the journal is the best option, just as a high APC does not prove quality.

Instead of producing one overall ranking, ask decisive yes-or-no questions: does the journal publish your study design, are there comparable papers from the last two years, is the review procedure transparent, does the licence satisfy funder requirements, and can you meet any mandatory fee? A journal that fails a critical condition can be removed regardless of its other advantages.

What to do after a desk rejection

An editorial rejection before peer review is disappointing, but it can save months. Separate the reasons into two groups. A mismatch with scope or editorial priorities usually means that the manuscript should be redirected rather than rewritten from scratch. A comment about unclear novelty, a weak abstract or confusing structure may justify revision before the next submission.

Do not send the same files to another journal an hour later. First check the new journal's reference style, abstract length, section order, ethics statements, data requirements and conflict-of-interest declarations. Changing the addressee in the cover letter is not the same as adapting a submission.

If an editor offers a transfer within a publishing group, evaluate the proposed journal exactly as you would any other candidate. A convenient transfer button does not prove that the journal suits your audience, budget or indexing needs.

What a response to a reviewer can look like

Reviewer comment: “It is unclear why participants younger than 18 years were excluded.”

Weak response: “Thank you. We believe this criterion is justified.”

Stronger response: “Thank you. The study uses a scale validated for adults, so participants younger than 18 years were not included. We have added this explanation to Methods, paragraph 2, and added the corresponding limitation to Discussion.”

The stronger version states the decision, the basis for it and the location of the revision. If you disagree with a reviewer, the structure remains the same: acknowledge the point, explain your reasoning and state whether the manuscript was changed for clarity. There is no need to argue with the reviewer's tone.

When your study does not fit standard IMRaD

Different article types follow different conventions. A systematic review, methodological paper, qualitative study, case report or protocol may have its own reporting structure. Follow the target journal's instructions and the reporting guidance for your discipline. IMRaD is useful as the logic of an original empirical study, not as a universal mould imposed on every manuscript.

Co-authors should agree before submission

Before uploading the manuscript, confirm author order, affiliations, the corresponding author and each contributor's role. These details become harder to change after acceptance, and authorship disputes should not begin when the journal is already requesting final forms.

Check conflicts of interest, funding, ethics committee approval and study registration where relevant. These are not decorative appendices. A missing mandatory statement can halt editorial processing regardless of the strength of the results.

If a journal asks you to suggest reviewers, nominate specialists with relevant expertise and no obvious conflict of interest. Do not suggest close collaborators, members of your own group or recent co-authors when the journal's policy excludes them. The editor makes the final selection; your list is assistance, not a way to control the outcome.

Before submission, ask one co-author to read only the title, abstract, tables and conclusion. If those four parts tell different versions of the study, the manuscript is not ready. Inconsistency often appears after several rounds of internal editing.

Where to explain the work to people outside your field

A scientific article is written for a scientific journal and its reviewers. After publication, or alongside the publication process where journal policy permits, you can prepare a separate popular-science piece that explains the question, method, finding and limitation without specialist jargon. It does not replace the scientific paper and should never pretend to be one.

What Columnist provides

The Mosaic network includes the popular-science publication Science Official. Through Columnist, a researcher can publish a signed author column that explains a study in plain language and appears on an author page. There is no scientific peer review in this format, so the column cannot substitute for a journal article. Its purpose is communication with readers outside a narrow specialty.

After your scientific submission, explain the work in accessible language through Columnist, at columnist.mib.news.

FAQ

All guides in the handbook