How to write each section of a research paper

Start with the section you are writing now

Choose a structure that fits the kind of research you are doing

Fictional learning example: let the task determine the route

This is a fictional learning example only. A student's course asks for a review of research on an invented campus policy. The student does not create a Results section because they did not collect data. Instead, they group sources by claimed benefits, implementation concerns, and evidence limits; the discussion explains what the comparison suggests and where reviewed evidence remains limited. A different empirical assignment could require Methods and Results instead.

Build the paper through linked decisions

1. Identify the problem, reader, task, and evidence expectations.
2. Use initial reading to refine a question, purpose, and workable structure.
3. Give each section one reader-facing job, then use headings and transitions to show the path.
4. Make the claim, methods or source approach, reported evidence, and interpretation agree with one another.
5. Revise recursively: test the whole argument, then sections, paragraphs, sentences, and citations.

Give each section a contract, then test the whole route

A section can look fluent in isolation but still fail the paper when its purpose, evidence, and conclusion no longer align. Treat each section as a reader-facing contract and reverse-check the completed draft from the research task through the final contribution.

1. Confirm the paper type. Re-read the prompt and identify whether you report an original study, review literature, make an argument, propose research, or document work in progress.
2. Write a section contract. For every heading, state the question it answers, the evidence or source record it uses, and what readers should understand before moving on.
3. Trace the evidence path. Check that the Introduction's question or claim is genuinely addressed by the source approach, method, analysis, reported material, and final explanation.
4. Preserve reporting and interpretation boundaries. When the assignment separates sections, keep factual results in Results and reserve explanations, comparisons, limits, and implications for Discussion.
5. Account for real variations. A literature review, multi-experiment report, in-progress project, or course assignment may need a different route; use a variation only when the task and evidence justify it.
6. Reconcile the completion layer. Draft Abstract and Conclusion after the paper is stable, then check that they summarize the same purpose, method or approach, key answer, and bounded contribution.

Use common structures carefully

George Mason University Writing Center describes IMRaD as a common route for planned, systematic research and separates Introduction's research case, Methods' procedural account, Results' reporting, and Discussion's interpretation. UC San Diego Psychology outlines a common APA experimental-report structure while noting variations for literature reviews, multiple experiments, incomplete research, and course assignments. UW–Madison Writing Center treats research writing as recursive and recommends organizing a body around the points you want to make rather than allowing sources to determine the order. Follow instructor, department, publisher, and disciplinary conventions whenever they differ.

Use each section for its own job

Strong papers keep their functions distinct: an introduction establishes purpose, methods explain how a study was carried out, results report findings, discussion interprets them, and an abstract concisely summarizes the completed paper. This structure helps readers follow an argument and makes revision more focused.