2.1 The IMRaD Structure¶
What you will learn on this page
- The standard structure used in many English research papers (IMRaD)
- The role of each section (Introduction, Methods, Results, and Discussion)
- Disciplinary variations in structure
- The role of the Abstract and how to write it
- Rough length guidelines for each section
What is IMRaD?¶
IMRaD is one of the most widely used structures for research papers written in English. The name is an acronym formed from the first letters of the four main sections.
- Introduction
- Methods
- Results
- and
- Discussion
This structure is standard not only in the natural sciences, but also in empirical research in the social sciences and applied linguistics.
The role of each section¶
Introduction¶
This section explains the background, purpose, and significance of the study. It helps readers understand why the study is needed.
- Background and overview of prior research in the field
- Identification of problems or gaps in prior research
- Statement of the study purpose and research question(s)
AI prompt example: a quick check of Introduction structure
Please check whether the following Introduction includes all three elements:
(1) background, (2) a gap in prior research, and (3) the purpose of the present study.
If any element is missing, point it out. Do not add new content.
[Paste the Introduction here]
For a more detailed check aligned with the CARS model (importance, prior work, gap, and purpose), see Step 3: Check the structure in:
3.2 Writing the Introduction and Background
Methods¶
To ensure reproducibility, this section describes what you did and how you did it, with enough detail for readers to understand and evaluate the study.
- Participants or data sources
- Materials and instruments
- Procedures and data collection
- Data analysis
Results¶
This section reports the findings objectively. Interpretation is not the main focus here.
- Results of statistical analyses
- References to tables and figures
- Answers to the research question(s)
Discussion¶
This section interprets the meaning of the results, connects them to prior research, and discusses implications and limitations.
- Interpretation of results and comparison with prior work
- Theoretical and practical implications
- Limitations of the study
- Directions for future research
Conclusion¶
In many journals, the conclusion is included at the end of the Discussion, but some papers include a separate Conclusion section.
- Brief summary of key findings (not a repetition of the Results, but a concise statement of what they mean)
- Practical and theoretical implications
- Future directions
Discussion vs. Conclusion
The Discussion interprets and develops the argument around the results. The Conclusion summarizes what can be said as a takeaway. Do not introduce new arguments in the Conclusion.
The hourglass model of IMRaD¶
The IMRaD structure is often visualized as an “hourglass.”
The Introduction moves from a broad context to specific research questions. Methods and Results provide the most concrete details. The Discussion then expands from interpretation back to broader implications.
Keeping this “general → specific → general” flow in mind helps the overall logic of the paper remain clear.
The role of the Abstract and how to write it¶
The Abstract appears at the beginning of the paper, but it is usually written last. It summarizes the whole study in about 150–300 words.
Elements to include in the Abstract¶
| Element | Content | Rough proportion |
|---|---|---|
| Background | 1–2 sentences of context | 10–15% |
| Purpose | The study purpose | 1 sentence |
| Methods | A brief overview of the methods | 20–25% |
| Results | The main findings | 30–35% |
| Conclusion | Conclusions and implications | 15–20% |
Structured Abstract vs. Unstructured Abstract¶
Some journals require a Structured Abstract with headings (Background / Methods / Results / Conclusions). Others require an Unstructured Abstract in paragraph form. Always follow the submission guidelines of your target journal.
AI prompt example: drafting an Abstract
Using the summaries of each section (Introduction, Methods, Results, Discussion) below,
write an Abstract of no more than 250 words.
Requirements:
- Follow this order: Background (1–2 sentences) → Purpose (1 sentence) → Methods → Results → Conclusion
- Do not add any information not in the manuscript
- Spell out abbreviations at first mention
[Paste the section summaries here]
In many fields, the Abstract compresses IMRaD into a single paragraph. To improve consistency, it is effective to check whether your Abstract includes the IMRaD elements.
AI prompt example: checking an Abstract (standalone)
Please check the following Abstract from these perspectives:
(1) Does it include all five elements: Background / Purpose / Methods / Results / Conclusion?
(2) Does it contain any information not found in the main text (new claims)?
(3) Is the word count within the guideline range (roughly within 250 words)?
If there are problems, point them out only. Do not propose a rewritten version.
[Paste the Abstract here]
For a detailed final check of your Abstract and title (numbers, claims, and conclusions), see:
3.5 Revision Techniques
Rough length guidelines for each section¶
The total length varies by journal, but for a typical empirical paper (about 6,000–8,000 words), a common guideline is:
| Section | Share of the whole | Rough guide for a 6,000-word paper |
|---|---|---|
| Introduction | about 20–25% | 1,200–1,500 words |
| Methods | about 15–20% | 900–1,200 words |
| Results | about 20–25% | 1,200–1,500 words |
| Discussion | about 25–30% | 1,500–1,800 words |
Disciplinary variations in structure¶
IMRaD is a common default, but disciplines often adapt it.
| Field | Common variation |
|---|---|
| Applied linguistics | Literature Review often appears as a separate section |
| Education | A Background / Context section may be included |
| Computer science | Related Work is often placed before Methods |
| Humanities | Papers may use argument-driven structures rather than IMRaD |
A common structure in computing and information science
In computing and computer science, the following structure is often seen.
- Introduction
- Related Work
- Proposed Method / System
- Experiments / Evaluation
- Results and Discussion (often combined)
- Conclusion
In this structure, “Proposed Method / System” corresponds to the Methods of IMRaD, but it often includes design rationale for systems or algorithms, not only procedures. Always check the template of your target journal or conference.
Confirm the conventions of your field
The most reliable approach is to read 3–5 recent papers from your target journal and identify the typical structure.
If you ask generative AI “What is the typical structure in this journal?”, it may only return a generic pattern.
In practice, using AI effectively requires familiarity with the papers in your own field.
A whole-structure check with AI¶
You can ask AI to check whether the overall structure of your paper is reasonable.
AI prompt example: a whole-structure check
Please check the following paper from the perspective of the IMRaD structure.
Checklist:
(1) Does each section fulfill its expected role?
(2) Are there overlaps or contradictions across sections?
(3) Is the flow Introduction → Methods → Results → Discussion coherent?
(Do the Results and Discussion actually answer the research questions stated in the Introduction?)
If you find problems, specify the section name and the concrete issue.
[Paste the manuscript here]
AI prompt example: mapping research questions to results