How to Write a Scientific Article Step by Step

Publishing scientific articles is a fundamental element of an academic career. Regardless of the field, the ability to prepare a well-written manuscript determines success in the world of science. A scientific article is not merely a research report — it is a communication to the scientific community that must be precise, logical, and convincing. The process of creating a scientific publication involves many stages: from research conception, through data collection and analysis, writing successive drafts, to navigating peer review and final publication. In this comprehensive guide, we discuss each of these stages, sharing practical tips based on experience working with researchers across various scientific disciplines.

Topic Selection and Formulating the Research Question

The first step in writing a scientific article is selecting a topic and formulating a precise research question. The topic should be current, relevant to the field, and feasible within available resources — both in terms of time and materials. It is worth conducting a preliminary literature review (a scoping review) to ensure the topic has not already been exhaustively covered, while confirming that sufficient literature exists to embed the study in a theoretical context.

A good research question should meet the FINER criteria: Feasible, Interesting, Novel, Ethical, and Relevant. Avoid questions that are too general ('How does X work?') — instead, formulate a specific, verifiable question ('Does the application of method X improve outcome Y in population Z?'). A precisely formulated research question determines the entire subsequent structure of the article, from method selection to results interpretation.

Before you start writing, create an outline of the article. List the main points you want to discuss in each section, key results you want to present, and main conclusions. Such an outline serves as a roadmap and helps maintain argument coherence throughout the text. Many experienced researchers recommend starting with the results section — when you know what you want to say, it's easier to build the rest of the article around that message.

Article Structure — The IMRaD Model

The structure of a scientific article is based on the IMRaD model (Introduction, Methods, Results, and Discussion), which is the standard in most scientific journals. This model ensures a logical flow of information from the research problem, through procedure description, data presentation, to their interpretation. Some journals modify this format — e.g., adding a separate 'Conclusion' section or combining 'Results and Discussion' into one section — but the fundamental logic remains the same.

The Introduction should guide the reader from the general context of the problem to the specific research question. Apply a 'funnel' structure: begin with the broad field context, narrow to the specific research problem, identify the gap in existing knowledge, and finally clearly state the study objective and hypotheses. The last paragraph of the introduction should explicitly state what the article contributes to existing knowledge.

The Methods section must be detailed enough for another researcher to replicate your study. Describe: the study population and inclusion/exclusion criteria, the research procedure step by step, tools and materials used (including software versions), data analysis methods, and ethical aspects (ethics committee approval, informed consent). A common mistake is writing methods too vaguely — 'a survey was conducted among students' is insufficient; provide the number of participants, recruitment method, research instrument, and data collection procedure.

Writing the Results Section

The Results section presents data objectively, without interpretation. Begin with a description of the research sample (demographic characteristics, variable distributions), then present results corresponding to each research question or hypothesis. Every statistical result should include the test statistic value, degrees of freedom (if applicable), p-value, and effect size measure.

Tables and figures are key elements of the results section. Follow this principle: data that can be presented in one sentence, describe in text; data requiring comparison of multiple values, place in a table; trends and distributions, present in graphs. Every table and figure must have a number, title, and be discussed in the text. Avoid repeating the same data in both text and table — the text should summarize and draw attention to the most important patterns, while the table presents detailed data.

Do not omit statistically non-significant results — these are important findings too. The tendency to publish only positive results (publication bias) is a serious problem in science. Honest presentation of all results builds the credibility of your study and article.

Discussion and Conclusions

The Discussion is where you interpret results in the context of existing literature and theory. Begin with a summary of the most important finding — what is the main discovery of your study? Then relate each significant result to previous research: do your results confirm or contradict existing findings? If results are inconsistent with the literature, propose possible explanations.

Honestly describe the study's limitations — every study has them. Limitations may concern sample size, participant recruitment method, research instrument, cultural or temporal conditions. Identifying limitations does not weaken the article — on the contrary, it demonstrates research maturity and enables other researchers to properly interpret your results. At the end of the discussion, present practical and theoretical implications of your findings and propose directions for future research.

Conclusions should be concise — 1-2 paragraphs summarizing the main discoveries and their significance. Avoid introducing new information in this section. Conclusions should directly answer the research question posed in the introduction, closing the logical arc of the article.

Literature Review and Citations

Literature review is a key element of every scientific article. Use reliable databases: PubMed and MEDLINE for biomedical sciences, Scopus and Web of Science for interdisciplinary research, Google Scholar as a supplement (though with caution — it also indexes non-peer-reviewed sources), and specialized databases for your field. A systematic literature review should cover the most important works from the last 5-10 years and classic publications that founded the field.

Cite sources according to the requirements of the chosen journal — most commonly in Vancouver style (numerical) in biomedical sciences or APA in social sciences and humanities. Use a bibliography manager (Zotero, Mendeley, EndNote) from the very start — manually managing citations with 40-100 references is a recipe for errors and inconsistencies. Ensure that every claim requiring support has an appropriate citation, but avoid overcitation, which clutters the text.

Pay special attention to self-citations. Citing your own previous work is acceptable and often necessary — especially when your study is a continuation of previous research. However, excessive self-citation (above 10-15% of all citations) may be viewed negatively by reviewers and editors. Always cite the most relevant works regardless of authorship.

Revision and Text Editing

After writing the first draft, spend time on multiple revisions. The revision process should proceed on several levels. Structural revision: does the article have a logical structure? Does each section flow smoothly to the next? Do conclusions follow from the presented data? Content revision: is the argumentation coherent? Do data in the text match data in tables and figures? Are all citations correct?

Language revision: are sentences clear and concise? Are there repetitions, unnecessary words, and jargon? Scientific writing style does not mean complicated writing — on the contrary, the best articles are characterized by simplicity and precision of language. Avoid passive voice where active voice is more natural. Instead of 'The study was conducted by the authors,' write 'We conducted the study.'

Ask co-authors or colleagues for critical reading (peer feedback) before submitting the article. Fresh eyes will catch errors and ambiguities that you missed after reading your own text multiple times. Professional language editing, especially for publications in English by non-native authors, significantly increases acceptance chances. Many journals reject articles due to poor language quality before they even reach substantive review.

Choosing the Right Journal

Choosing the right journal is a strategic decision worth making even before you start writing. Consider: thematic scope — does your article fit the journal's profile? Impact factor and other metrics (CiteScore, SJR) — a higher IF doesn't always mean a better choice; what matters is thematic fit. Review time — some journals process articles in 2-3 months, others in 6-12. Open access policy — OA publication increases reach but may involve fees (APC). Rejection rate — top journals reject 80-95% of submissions.

Prepare a cover letter that briefly presents the significance of your research, justifies your choice of journal, and confirms that the work is not simultaneously under consideration elsewhere. A good cover letter is 3-4 paragraphs: context and problem, what your study contributes, why this journal, and confirmation of originality and all co-authors' consent.

The Peer Review Process and Responding to Comments

After submission, the peer review process begins. In the single-blind model, reviewers know the authors' identities but authors don't know the reviewers. In the double-blind model, both sides are anonymous. Open peer review, where identities are public and reviews are published alongside the article, is becoming increasingly popular.

Receiving reviewer comments is a critical moment. Even if the comments seem harsh, remember that the goal of review is to improve the article's quality. Prepare a detailed response letter addressing each comment point by point. For each comment: quote it, describe what changes you made in the article (or justify why you disagree), and indicate the exact location in the text where changes were made. A polite, factual tone in your response, even to critical comments, significantly increases the chances of a positive editorial decision.

Typical editorial decisions after review are: accept (acceptance without changes — rare), minor revision (small corrections), major revision (significant changes and re-review), reject with resubmission (rejection with the possibility of resubmission after thorough changes), and reject (rejection). A 'major revision' decision is not a failure — on the contrary, it means reviewers see potential in your work. The most important thing is to thoroughly address all comments and implement the required changes.

Publication Ethics and Authorship

Publication ethics is the foundation of credible science. Principles set by ICMJE (International Committee of Medical Journal Editors) and COPE (Committee on Publication Ethics) include: honest reporting of results (do not fabricate or falsify data), proper attribution of authorship (an author must have a substantial contribution to conception, execution, or interpretation of the research), avoiding plagiarism (including self-plagiarism — republishing your own results without proper citation), and declaring conflicts of interest.

The question of authorship can be a source of controversy in research teams. Agree on author order at an early stage of the project. In many fields, the first author is the person who did most of the work, and the last is the project leader (senior author). All persons listed as authors must meet ICMJE criteria: substantial contribution to the conception or execution of the study, participation in writing or critical revision of the text, approval of the final version, and agreement to bear responsibility for the article's content. Persons who helped but don't meet these criteria should be listed in the Acknowledgments section.

If you need support with manuscript preparation, explore our comprehensive publication support services. You may also find our guide on editing and proofreading scientific texts helpful. Have questions? Contact us — we respond within 24 hours.