Reviewers do not read research papers the way readers read novels, front to back and open to surprise. They navigate by expectation: methods should be here, the key result there, the limitations two paragraphs into the discussion. The IMRaD structure, short for Introduction, Methods, Results, and Discussion, is the shared map that makes this navigation possible. Papers that follow the map get evaluated on their substance; papers that ignore it force reviewers to hunt, and hunting reviewers grow impatient long before they grow generous.
This guide explains what belongs in each section, why the format earns its dominance in empirical fields, how to adapt it when your work is qualitative or interpretive, and the section-bleed mistakes that most often send manuscripts back for revision.
What the IMRaD Structure Is
IMRaD names the four core sections of an empirical research paper: an Introduction that establishes the problem and question, a Methods section that documents how the work was done, a Results section that reports what was found, and a Discussion that interprets those findings. The lowercase a stands for and. Over the twentieth century the format became the default across medicine, the natural sciences, and much of the social sciences, and most journals in those fields now expect it either explicitly in author guidelines or implicitly through reviewer habits.
A helpful mental image is the hourglass. The introduction starts wide with the field’s concern and narrows to your specific question. Methods and results sit in the narrow middle, tightly focused on your study. The discussion widens back out, connecting your findings to the larger conversation the introduction opened.
Why Reviewers Reward the Format
The IMRaD structure is not an aesthetic preference; it is a division of labor. Each section answers exactly one reader question, which lets a reviewer verify the logic of a study in a single pass: was the question worth asking, was the method capable of answering it, do the reported results actually follow, and does the interpretation stay within what the evidence supports. When material appears in the wrong section, that verification chain breaks. Interpretation smuggled into results makes findings look inflated; results previewed in methods confuse the sequence of evidence; a discussion that introduces new data reads as an appendix in disguise.
The IMRaD Structure, Section by Section
Introduction: From Territory to Question
A working introduction makes three moves in order: establish the territory and why it matters, identify the gap or tension in what is known, and state your question or purpose, often with a brief preview of the approach. Aim for the narrowest credible opening; beginning three levels of generality above your topic wastes the reader’s patience. If the question itself still feels soft, sharpen it first using the process in our guide to writing a research question, because no structure can rescue a paper built on a vague one.
Methods: Enough to Judge and Repeat
The methods section serves two audiences: reviewers deciding whether your approach can support your claims, and future researchers who might replicate or extend the work. Report design, participants or materials, procedures, and analysis in the past tense, with enough specificity that a competent colleague could repeat the study. Justify the choices a skeptic would question, such as sample size or instrument selection, and save the defense of your findings for the discussion.
Results: Report Without Interpreting
Results answer one question only: what did you find? Present findings in an order that mirrors your research questions, use tables and figures for detail while the text states the pattern, and resist explanatory language. The sentence participants in the flipped section scored higher belongs here; the sentence this suggests active learning drove the difference belongs one section later. That discipline feels artificial at first and quickly becomes the easiest part of the format.
Discussion: Interpret, Bound, and Connect
Open the discussion by answering your research question directly in one or two sentences, then interpret: how the findings relate to prior work, what might explain them, and what they mean for practice or theory. State limitations honestly and specifically; reviewers read vague limitation paragraphs as evasion and precise ones as competence. Close with implications and a future-work paragraph that is more than ritual, naming the next study someone could actually run.
One Study, Four Opening Sentences
Register shifts between sections are easier to hear than to describe, so consider a hypothetical study of a flipped statistics course. The introduction might open: while flipped designs are widely adopted in quantitative courses, evidence from open-enrollment institutions remains thin. The methods: we compared two sections of introductory statistics at a community college, one flipped and one lecture-based, across a single semester. The results: the flipped section averaged five points higher on the common final, with the largest gap on interpretation items. The discussion: these findings suggest the benefit of flipping concentrates where students must explain rather than compute. Four sentences, four distinct jobs: justify, document, report, interpret. If you can draft those four sentences for your own project, the paper has a spine, and each section simply elaborates its sentence.
A Paragraph-Level Map
| Section | Reader’s Question | Typical Share of the Paper |
|---|---|---|
| Introduction | Why should I care, and what did you ask? | Roughly 15 to 20 percent |
| Methods | How did you find out, and can I trust it? | Roughly 20 to 25 percent |
| Results | What did you find? | Roughly 25 to 30 percent |
| Discussion | What does it mean, and what are its limits? | Roughly 25 to 30 percent |
These shares are conventions, not rules; journals and fields vary. Their value is diagnostic: a manuscript whose introduction consumes a third of the pages, or whose discussion is two paragraphs, usually has a balance problem worth fixing before submission.
Using the IMRaD Structure Beyond the Lab Sciences
Qualitative researchers adapt the format routinely: methods becomes a fuller account of positionality, data collection, and analytic approach, while results and discussion sometimes merge into a findings section that weaves evidence and interpretation, where the journal allows it. Humanities scholarship built on argument rather than data generally should not force itself into IMRaD at all; essays in history or literary studies follow a thesis-driven architecture with its own logic. The transferable lesson is the contract behind the format, not the headings: tell readers what question you are pursuing, show them your evidence and how you gathered it, and keep claims proportional to that evidence.
From Dissertation Chapter to IMRaD Article
Many early-career researchers carve their first article out of a thesis, and the conversion is structural rather than cosmetic. A dissertation literature review surveys the field to demonstrate competence to a committee; an article introduction builds exactly one gap in a handful of paragraphs, citing only what that gap requires. One article answers one question, which means a three-question dissertation is usually three papers rather than one long one. Methods shrink to what the chosen question needs, with full instruments and protocols moved to appendices or an online repository. The discussion sheds its defensive, examination-day hedging in favor of a confident, bounded interpretation.
A practical exercise makes the compression concrete: before reopening the dissertation file, write the four opening sentences described above for the article you intend, then import only the material each sentence needs. Authors who start from the old document instead tend to produce an abridged thesis, and reviewers can always tell.
Common IMRaD Mistakes Reviewers Flag
- Section bleed. Interpretation in results, results in methods, new evidence in the discussion. Label each paragraph with the question it answers and relocate the strays.
- The overstuffed introduction. A literature review that catalogs everything instead of building the gap your question fills.
- Methods amnesia. Missing details a replicator would need, especially about analysis decisions and exclusions.
- The mirror-image discussion. A discussion that merely restates results in different words rather than interpreting them.
- Abstract mismatch. An abstract promising what the sections do not deliver; write it last, from the finished paper.
One further habit prevents most bleed before it starts: terminology discipline. Call the intervention, the groups, and the measures by the same names in all four sections. If the introduction promises to study engagement and the results report participation scores, reviewers are left wondering whether those are the same construct or a quiet substitution. A final find-and-replace pass hunting for drifting synonyms is cheap insurance against an avoidable revision round.
Structure solves the architecture; the sentences still need work. The habits in our collection of academic writing tips pair well with a sound skeleton.
Where IMRaD Meets the Conference Circuit
Conference work follows the same skeleton at smaller scale: a proposal is an IMRaD paper in miniature, and a fifteen-minute talk is one with the methods compressed and the so-what expanded. That continuity is why presenting early drafts pays off, since reviewer and audience questions reveal which section is underbuilt while revision is still cheap. Papers accepted at NIVA’s virtual conferences pass double-blind review and appear in the ISBN conference proceedings, and the strongest of them often mature into submissions for NIVA’s journals, carrying the same structure through each stage.
Frequently Asked Questions
Each section of IMRaD has its own logic and its own failure modes. For the first, see how to write the introduction of a research paper.
IMRaD stands for Introduction, Methods, Results, and Discussion, the four core sections of most empirical research papers. The lowercase a simply stands for and. Journals may add an abstract, references, and appendices around that core.
No. It dominates medicine, the natural sciences, and much of the social sciences, but humanities essays follow thesis-driven structures, and some qualitative journals allow merged findings sections. Always check the target venue’s author guidelines first.
Results report what you found, in text, tables, and figures, without explanation. Discussion interprets those findings: what they mean, how they relate to prior work, what limits them, and what should happen next. Keeping the two separate protects your credibility.
Most writers draft methods and results first, since they are concrete and already settled, then the discussion, then the introduction, and the abstract last. Drafting in that order prevents the introduction from promising a paper you did not write.
Got a study organized and a question answered? Give the structure its first audience: apply to present at NIVA’s upcoming virtual conference. Faculty registration is $150, students $70, membership costs nothing, and accepted papers join the ISBN proceedings.