How to write a scientific paper — a practical guide

How do you write a scientific paper?

Writing a scientific paper is not only presenting data—it is answering a research question in a systematic, transparent, and persuasive way. A strong paper is methodologically sound, clearly written, and aligned with the literature.

This guide walks through the main stages from research idea to publication.

1️⃣ From research idea to study design

1.1 Research question

A strong paper starts with a strong question.

• Specific
• Measurable
• Addresses a gap in the literature
• Delivers scientific or clinical value

Weak question: “A review of hydrocephalus”

Stronger question: “Which independent clinical factors predict ETV success in premature infants?”

1.2 Literature review

Goals:

• What has been done?
• Where are the contradictions?
• What are methodological gaps?
• What will your contribution be?

Search databases such as PubMed, Scopus, and Web of Science systematically. Combine keywords and MeSH terms where appropriate.

1.3 Study design

The question determines the design.

If the design is wrong, the paper will be weak regardless of prose.

2️⃣ IMRAD structure

Most scientific articles follow IMRAD: Introduction, Methods, Results, And Discussion.

This builds a logical story: Problem → How I studied it → What I found → What it means.

3️⃣ Section-by-section writing

🧠 Introduction

Aims:

• Define the problem
• Show the literature gap
• State the study purpose clearly

Typical flow:

• Broad context
• What is known
• What is missing
• Objective of this study

The last sentence often states: “The aim of this study was to …”

🔬 Methods

The most critical section.

Should include:

• Study design
• Setting and timeframe
• Inclusion/exclusion criteria
• Variables measured
• Measurement procedures
• Statistical analysis

Core principle: another researcher should be able to repeat the study.

Statistics should report:

• Test names
• Significance level
• Software
• Assumption checks where relevant

📊 Results

Present data without interpretation. No discussion here.

Suggested order:

• Participant characteristics
• Primary analysis
• Secondary analyses
• Regression / subgroup analyses

Tables and figures should complement—not repeat—the text verbatim.

💡 Discussion

Here you:

• Summarize key findings
• Compare with prior literature
• Discuss mechanisms where appropriate
• Highlight clinical or scientific implications
• State limitations openly

Strong papers:

• Do not hide limitations
• Are cautious with causal language when design does not support it
• Avoid overgeneralization

📎 Conclusion

Keep it short and concrete.

• Summary of findings
• Clinical/academic implication
• No unnecessary repetition

4️⃣ Title and abstract

Title

• Concise
• Specific
• Not overstated
• Avoid implying causality inappropriate to the design

Weak example: “ETV is the best treatment…”

Better example: “Factors associated with ETV success in premature infants”

Abstract

Often structured:

• Objective
• Methods
• Results
• Conclusion

Common mistake: discussing implications without presenting data.

5️⃣ Publication ethics

Scientific writing must avoid:

• Plagiarism
• Fabricated data
• Selective reporting
• Authorship that does not reflect real contribution

Ethics committee information should be stated clearly when required.

6️⃣ Journal choice and submission

Consider:

• Scope
• Target readership
• Recent publications
• Methodological expectations

Cover letter:

• Summarize contribution in 4–5 sentences
• Stay factual—avoid hype

7️⃣ Peer review and revision

When revisions arrive:

• Respond professionally, not emotionally
• Answer every point

Response document should map:

• Reviewer comment
• Your reply
• Where the manuscript changed (line/page)

8️⃣ Common mistakes

• Strong discussion, weak methods
• Over-focusing on p-values
• Misusing “predictor” language in retrospective designs
• Omitting limitations
• Table/text contradictions

Conclusion

A strong scientific paper:

• Starts from the right question
• Rests on sound methods
• Presents results honestly
• Relates fairly to the literature
• Draws measured conclusions

Scientific writing is as much intellectual discipline as technique.