Research Paper Format: IMRAD Structure Explained with Examples

Introduction

You finished your data collection, opened a Word document, and froze. Do you start with the methods? The literature? The results? The introduction? Without a clear format, even good research becomes a confused paper. The good news is that nearly every empirical journal in the world uses the same structure — IMRAD — and once you understand its logic, writing a paper stops being a creative puzzle and becomes a predictable sequence.

IMRAD stands for Introduction, Methods, Results, And Discussion. The acronym was formalised by the American National Standards Institute in 1972 and adopted by the ICMJE in the late 1970s; today, more than 90% of empirical articles in Scopus-indexed journals follow it. Understanding IMRAD structure — and why each section exists — is the single most useful research paper format you can learn.

This guide walks through each IMRAD section with an annotated example, the common mistakes that get papers desk-rejected, and three structural variations you will encounter in journals that deviate from strict IMRAD. If you are also drafting the full paper, our research paper writing service works alongside you from outline to submission.

What Is IMRAD?

IMRAD is a rhetorical structure, not a content rule. It tells you in what order to present information, not what information to collect. Each section answers a different question for the reader:

• Introduction: Why did you do this study?
• Methods: How did you do it?
• Results: What did you find?
• Discussion: What does it mean?

Notice the symmetry: the Introduction sets up a question, the Discussion interprets the answer; the Methods describes the procedure, the Results report the output. A well-written IMRAD paper reads as a closed loop — the introduction’s promise is fulfilled by the discussion.

IMRAD is not a cage. It is a contract between writer and reader about where each piece of information will be found.

The Introduction: Opening Moves

Most strong introductions follow the CARS model (Create A Research Space) developed by John Swales in 1990 and refined in 2004. CARS has three moves:

1. Move 1 — Establish a territory: state the broad importance of the area. Cite recent reviews to show the field is active.
2. Move 2 — Establish a niche: identify a gap, contradiction, or unaddressed extension. Words like “however,” “despite,” “little is known” signal the gap.
3. Move 3 — Occupy the niche: state explicitly what your paper does to fill the gap, ending with the aim, hypothesis, or research question.

A typical Introduction runs 600–1,200 words and cites 15–40 sources, narrowing from broad context to the specific question. Avoid the common mistake of writing a literature review here — the deep review belongs in a standalone section if the journal requires one (common in social science) or is integrated into Move 2. For a fuller treatment, see our thesis introduction writing guide, which applies the same CARS logic.

End the Introduction with a clear statement of purpose. The sentence “This study aimed to examine the effect of X on Y in population Z” tells the reader exactly what to expect in Methods and Results.

The Methods: Reproducibility Is the Goal

The Methods section is judged by one criterion: could another researcher replicate your study from the information given? If yes, the section works. If no, it does not. A useful mental exercise is to imagine a sceptical reviewer in another country trying to repeat your study using only this section.

Use subheadings to make the section scannable. Typical subheadings for a quantitative paper:

• Design: experimental, quasi-experimental, correlational, cross-sectional, longitudinal.
• Participants: sampling method, sample size with power calculation, inclusion/exclusion criteria, demographics.
• Materials/Instruments: what you used, with psychometric properties (reliability, validity) for standardised instruments.
• Procedure: step-by-step protocol, in the order performed.
• Data analysis: software (SPSS, R, NVivo), statistical tests, significance threshold, assumption checks.
• Ethics: IRB approval number, informed consent procedure, data anonymisation.

For qualitative papers, replace the statistical machinery with: philosophical paradigm, analytic framework (e.g., Braun and Clarke’s thematic analysis), coding approach, and trustworthiness criteria. Our research methodology guide covers both paths in depth.

The Results: Show, Don’t Interpret

The Results section reports what you found, without explaining what it means. Interpretation belongs in the Discussion. Mixing the two confuses readers and triggers reviewer pushback. A clean Results section follows three rules:

1. Order results to mirror your research questions or hypotheses. If your first hypothesis is about gender differences, report gender differences first.
2. Report descriptive statistics before inferential statistics. Means, standard deviations, and sample sizes come before t-tests, ANOVAs, or regressions.
3. Use tables and figures for patterns, text for specific values. A table is better than a paragraph listing 20 numbers; a figure is better than a table for showing a trend.

Follow APA reporting conventions: italicise statistical symbols, report exact p-values to three decimal places (p = .013, not p < .05), and always report effect sizes alongside p-values. Example: “An independent samples t-test showed that the intervention group scored significantly higher than controls, t(118) = 3.42, p = .001, d = 0.62, 95% CI [0.26, 0.98].”

Every table and figure must be self-contained — a reader should understand it without reading the surrounding text. Number tables and figures sequentially (Table 1, Table 2; Figure 1, Figure 2), give each a clear title, label every axis with units, and add notes for abbreviations. For a deep dive on statistical reporting, see our statistical tests in research guide and our walkthrough on interpreting SPSS output.

The Discussion: Make Sense of Findings

The Discussion is where you interpret, contextualise, and contribute. A common mistake is to simply repeat the Results in different words — that adds nothing. A strong Discussion follows Holliday’s five-move structure:

1. Summarise key findings: one paragraph restating the main results in plain language (no statistics).
2. Relate to literature: compare your findings to prior studies. Where do you agree, disagree, or extend? Cite specific studies.
3. State the contribution: what does your paper add that did not exist before? Be explicit — “This study is the first to…” or “These findings extend prior work by…”
4. Acknowledge limitations: sample size, generalisability, self-report bias, cross-sectional design. Honesty builds credibility.
5. Implications and future research: what should practitioners, policymakers, or researchers do next?

End with a brief conclusion paragraph (some journals want this as a separate section) that reiterates the contribution in 3–5 sentences. For more on the discussion and conclusion, see our guide on writing discussion and conclusion.

Annotated IMRAD Example

Here is a stripped-down IMRAD example for a study on flipped classrooms in Indian engineering colleges. Notice how each section does only its job.

Introduction (excerpt)

“Engineering education in India enrols over 2.5 million students annually, yet employer surveys consistently report low graduate employability (Move 1). Flipped classroom pedagogy has shown promise in Western contexts, but evidence from large Indian engineering cohorts remains limited (Move 2). This study examined whether a 12-week flipped classroom intervention improved academic performance and engagement among first-year engineering students at a private university in Karnataka (Move 3).”

Methods (excerpt)

“A quasi-experimental design was used with two intact sections (n = 142) taught by the same instructor. The intervention section (n = 71) received flipped instruction; the control section (n = 71) received traditional lectures. Pre- and post-tests assessed conceptual understanding; the 32-item Student Engagement Instrument measured engagement. Independent t-tests and mixed ANOVA were conducted in SPSS 28, with alpha set at .05.”

Results (excerpt)

“The intervention group showed significantly higher post-test scores (M = 78.4, SD = 9.2) than the control group (M = 71.1, SD = 10.8), t(140) = 4.32, p < .001, d = 0.73. Engagement scores were also higher in the intervention group (M = 4.12, SD = 0.42) compared to control (M = 3.78, SD = 0.55), t(140) = 4.18, p < .001, d = 0.70.”

Discussion (excerpt)

“These findings align with Lage et al. (2000) and extend them to a previously unstudied Indian engineering context. The effect size (d = 0.73) is medium-to-large by Cohen’s conventions, suggesting practical, not just statistical, significance. Limitations include a single-institution sample and reliance on self-reported engagement. Future research should test the intervention across multiple institutions and include qualitative data on student experience.”

IMRAD Variations: AIMRaD, IRDaM, and Beyond

Not every journal uses strict IMRAD. Common variations you will encounter:

• AIMRaD: Abstract + IMRAD. The default in most life-science and medical journals.
• AIM(RaD)C: Abstract, Introduction, Methods, Results and Discussion combined, Conclusion. Used in some engineering journals.
• IRDaM: Introduction, Results, Discussion, Methods (Methods at the end). Common in some chemistry and physics journals where the experimental detail is routine.
• OCAR: Opening, Challenge, Action, Resolution. Used in some earth-science and ecology journals; suggested by Schimel’s Writing Science.

Always check the journal’s author guidelines before drafting. Adapting a pre-existing manuscript to a different format takes 2–4 hours; writing from scratch in the wrong format takes far longer. Pair this guide with our journal selection guide and the broader how to write a research paper walkthrough.

Conclusion

IMRAD works because it matches the way science actually proceeds: question, method, finding, interpretation. Once you internalise this structure, writing a research paper becomes a matter of filling each section with the right kind of content rather than staring at a blank page. The format is not a constraint — it is a scaffold that lets your readers find what they need, when they need it.

If you want help applying IMRAD to your specific study, our research paper writing service supports each section individually or the whole manuscript. For the rest of the Research Paper series, browse our blog or contact us directly. Pair this guide with our pieces on abstract writing and citation styles to round out your publishing toolkit, and check our FAQ page for common formatting questions.

Frequently Asked Questions