A doctoral student sits in front of a blank page headed “1. Introduction” for three weeks. He writes an opening paragraph about the importance of digital transformation, deletes it, rewrites it and deletes it again. The problem is not his writing but his sequence: the Introduction is the hardest section to write when you do not yet know exactly what you will argue at the end.
IMRaD — Introduction, Methods, Results and Discussion — is the standard frame for empirical papers across the natural sciences, biomedicine, engineering and most social sciences. It exists because readers need to find information quickly: skimmers jump to the results tables, replicators jump to the methods. Once you understand that logic, each section becomes easier to write because you know what job it has to do.
One question per section
| Section | Reader’s question | Suggested share of words | Dominant tense |
|---|---|---|---|
| Introduction | Why was this study needed? | 10–15% | Present for established knowledge; past or present perfect for prior studies |
| Methods | What exactly was done, and could I repeat it? | 20–25% | Past |
| Results | What was found? | 20–25% | Past; present when referring to tables and figures |
| Discussion | What does it mean, and where are the limits? | 25–35% | Present for interpretation; past when restating findings |
These shares assume a paper of roughly 5,000–7,000 words and are guidance, not rules. But if your Introduction takes up 30% of the paper, it is almost certainly doing the work of a review article.
The hourglass
Picture the paper as an hourglass. The Introduction narrows from broad to specific: general context, what is known, the gap, your precise question. Methods and Results form the narrow waist, describing only your study. The Discussion widens again: main finding, comparison with other work, implications, limitations, next steps. A common error is a Discussion that widens in its first sentence without restating the key result, or an Introduction so narrow from the start that readers outside the subfield cannot follow.
The order in which to actually write
- Tables and figures: they are the backbone. Decide which results deserve a table and which need a figure.
- Methods: the easiest section, because you are describing what you did. Write it while the details are fresh.
- Results: guide the reader through tables and figures in the order that answers your research questions.
- Discussion: now you know what you found and why it matters.
- Introduction: write it to set up exactly what the Discussion will say. Every study you engage with in the Discussion should be planted in the Introduction.
- Abstract and title: last, when you know precisely what the paper says.
Drafted in this order, the Introduction and Discussion fit together as two halves of one argument rather than two independent essays.
What belongs nowhere near a given section
- Results inside Methods: “312 people took part, 68% of whom were women” usually belongs in Results as a sample description, unless the journal says otherwise.
- Interpretation inside Results: “This suggests teachers lack motivation” is interpretation and belongs in the Discussion. Results report what was observed.
- New results inside the Discussion: every number in the Discussion should already have appeared in Results.
- A long literature review inside the Introduction: a paper needs enough literature to establish the gap, not a retelling of the whole field.
Variations by field and article type
Not every paper is IMRaD. Systematic reviews follow their own structure guided by PRISMA. Theoretical and humanities papers are organised by argument rather than procedure. Many social science journals add a separate “Literature Review” or “Theoretical Framework” section between Introduction and Methods. Some biomedical journals place Methods at the end. Always read the author guidelines and three recent articles from your target journal before setting up your outline.
A one-page outline to fill before you draft
Before you write a single sentence, complete this:
- Research question (one sentence).
- Gap: what existing work has not answered (one sentence plus two or three key sources).
- Methods: design, sample, instruments, analysis (four bullet points).
- Three main findings, each tied to a table or figure.
- One-sentence contribution: what readers will know after reading that they did not know before.
- Two most important limitations.
If you cannot fill the contribution line, the paper is not ready to be written; go back to the data or talk to your co-authors.
Typical beginner mistakes with IMRaD
- Thin Methods: “a questionnaire was used” without naming the scales, the number of items or how scores were computed.
- Results that restate every number in the tables as prose.
- A Discussion that opens by re-reviewing the literature instead of with the main finding.
- Improvised headings such as “Chapter 3: Current Situation” in a journal article.
Do this now: paste the one-page outline at the top of your manuscript file, fill it in and begin with your tables and figures instead of your opening sentence. Once the Discussion has a rough draft, return to the Introduction — you will find it comes out shorter and sharper than your first attempt.
Câu hỏi thường gặp
What does IMRaD stand for?
IMRaD stands for Introduction, Methods, Results and Discussion, the standard structure of empirical research papers. Each section answers a distinct reader question: why, how, what was found and what it means.
Which section of a research paper should I write first?
Many experienced authors start with tables and figures, then Methods, Results and Discussion, and only then the Introduction, abstract and title. This order keeps the Introduction aligned with what the Discussion argues.
Can I interpret findings in the Results section?
Keep interpretation to a minimum. Results report what was observed; interpretation, comparison with other studies and implications belong in the Discussion, unless the journal uses a combined section.
Do social science papers use IMRaD?
Most empirical social science papers do, often adding a literature review or theoretical framework section. Theoretical and humanities papers are usually organised by argument instead.
How long should each section of a research paper be?
For a 5,000–7,000-word paper, a rough guide is 10–15% Introduction, 20–25% each for Methods and Results, and 25–35% Discussion. Check recent articles in your target journal.
