How to write a research paper

Before you write: the main finding and the target journal

Two decisions come before the first sentence.

First, write down the paper’s single main finding in two sentences. Not the topic, the finding. “We studied the adoption of UPI among small retailers in Coimbatore” is a topic. “Among 412 small retailers in Coimbatore, perceived transaction cost predicted UPI adoption more strongly than perceived security, reversing the pattern reported in urban studies” is a finding. If you cannot write the second kind of sentence yet, you are not ready to draft. You may need more analysis, or you may be trying to put two papers into one.

A PhD thesis usually yields two to four papers. Each one needs its own main finding. Pulling a whole chapter into a paper is the most common reason first drafts run to 12,000 words and read like a report.

Second, choose the journal before you draft. The journal decides the word limit, the reference style, whether the abstract is structured, how long the introduction can be and who the readers are. A paper written for “any Scopus journal” ends up fitting none of them. Our guide on how to choose a journal sets out a shortlisting method. Once you have picked one, download two or three recent articles from it that are close to your topic and keep them open while you write. They show you, better than any guideline, what that journal accepts.

The IMRaD structure and what each section does

IMRaD is not a formality. Each section answers one question, and readers jump straight to the section that answers theirs. A reviewer checking your sample size goes to Methods. A reader deciding whether to cite you goes to the abstract and the first paragraph of the Discussion.

The word counts above assume a 6,000-word limit, which is common in management and social science. Engineering and life-science journals often allow 3,500 to 5,000 words, and letters or short communications far less. Scale the proportions, not the content.

Some journals merge Results and Discussion into one section. Some put Methods at the end. Some add a separate Literature Review or Theoretical Framework section between the Introduction and Methods, especially in management and education. Follow the journal, and keep the underlying logic: problem, method, evidence, meaning.

The order in which to write a research paper

You read a paper from title to references. You should not write it in that order. The introduction is hardest to write first because you do not yet know precisely what the paper will claim. Most experienced authors work from the evidence outwards:

1. 1

   Tables and figures

   Decide which three to six displays carry your findings. The paper is built around them.

2. 2

   Methods

   The easiest section to write because you already know what you did. Write it while details are fresh.

3. 3

   Results

   Describe the tables and figures in a logical order, without interpreting them yet.

4. 4

   Discussion

   Interpret the main findings, compare with earlier studies, state limitations.

5. 5

   Introduction

   Now you know exactly what the paper shows, so you can set up the gap it fills.

6. 6

   Abstract and title

   Written last, from the finished paper, so they describe what is actually there.

Start with the tables and figures because they force decisions. Which comparisons matter? Which ones are supporting detail for a supplementary file? If you have 14 tables, you have either two papers or a thesis chapter.

Write the first draft quickly and roughly. Correcting a rough full draft is much easier than polishing the introduction for a month and never reaching the discussion.

What to put in each section of a research paper

Title

Aim for 10 to 20 words that name the main variables, the population or material, and where helpful the design. “Effect of fly ash replacement on the compressive strength of M30 concrete: an experimental study” works. “A study on concrete” does not. Avoid abbreviations and words that add nothing, such as “A study of” or “An investigation into”.

Introduction

Think of it as a funnel with three or four paragraphs (our introduction guide walks through the CARS pattern with an annotated example). Open with the broad problem and why it matters, with a citation or a figure. Narrow to what is already known about your specific question. Then say what is not known, and why that matters. The last paragraph states your aim or research question and, in many fields, briefly how you approached it. Keep the literature focused on the gap. A two-page history of the field belongs in your thesis, not here. If you find it hard to state the gap in a sentence, write it as a single line in your notes before drafting anything else, and use the same line later when you write the abstract.

Methods

Give enough detail for someone to repeat the study. That usually means study design; setting and dates; sample and how it was selected; materials, instruments or questionnaires and their source; procedure; and the analysis, including the software and version (SPSS 29, R 4.4, MATLAB R2024b and so on). Justify choices a reviewer might question, such as why you used a convenience sample or a particular statistical test. Our guide on choosing a statistical test helps if you are unsure. State ethics approval and consent for research with people or animals.

Many fields have reporting guidelines that journals expect you to follow: CONSORT for randomised trials, STROBE for observational studies, PRISMA for systematic reviews, among others. The EQUATOR Network keeps the full list. If one applies to your design, use its checklist while you write Methods, not after.

Results

Report what you found, in the same order as your objectives or hypotheses. Lead with the most important result. Use the text to point out patterns and give the key numbers; do not repeat every cell of a table. Report effect sizes and confidence intervals or standard deviations, not just p-values. Keep interpretation out. “Group A scored higher than Group B (mean difference 4.2, 95% CI 1.8 to 6.6)” belongs in Results. “This suggests that training improves confidence” belongs in Discussion.

Discussion

Open with a short paragraph restating the main finding in plain words. Then compare it with earlier studies: where you agree, where you differ, and a reasoned explanation for the differences. Follow with implications for theory or practice, then limitations, then directions for future work. Be careful with claims. Cross-sectional survey data cannot show causation, however much you want it to. Reviewers in every field look for overclaiming first.

Conclusion

Many journals want a short separate conclusion. Say what the study found and why it matters, in a few sentences. Introduce no new results and no new citations.

References

Cite what you actually read, not what another paper cited. Check every reference against the original: authors, year, volume, pages and DOI. Use a reference manager from the first draft; converting 60 references by hand into a new journal style after a rejection costs a full day.

Language and style for scholars writing in English as a second language

Reviewers are not looking for elegant English. They are looking for clear English. A few habits make most of the difference:

• One idea per sentence. Long sentences with several clauses are where grammar errors and ambiguity creep in. Split them.
• Use the same term for the same thing throughout. If you call it “customer loyalty” in the introduction, do not call it “consumer retention” in the results. Readers assume a different word means a different concept.
• Tense. Past tense for what you did and found (“we measured”, “scores increased”). Present tense for established knowledge and for what your tables show (“Table 2 shows”).
• Avoid inflated phrases. “It is observed that there is an increase in” becomes “X increased”. Indian academic English often carries formal padding from report-writing traditions; journal editors cut it anyway.
• Active voice is acceptable in most journals now, including “we”. Check recent articles in your target journal.

Read your draft aloud once. Sentences you stumble over are the ones a reviewer will stumble over too.

On AI writing tools

Most major publishers now allow limited use of AI tools for language editing but require you to disclose it, and none accept an AI tool as an author. Policies differ and are still changing, so read your target journal’s current policy. What no policy allows is text or references you have not checked. AI tools regularly invent citations that look real, and a single fabricated reference can sink a paper.

Revising and checking before you submit

Leave the finished draft alone for a few days, then read it as a reviewer would. Better still, give it to a colleague outside your immediate topic. If they cannot tell you the main finding after reading the abstract and introduction, the paper is not yet clear.

• The research question or objective appears in the last paragraph of the introduction, in one clear sentence.
• Every table and figure is cited in the text, in number order, and can be understood without reading the text.
• Numbers agree everywhere: abstract, results text, tables and figure labels.
• Every claim in the discussion points back to a specific result.
• Limitations are stated plainly, and you have said how they affect the conclusions.
• Every in-text citation is in the reference list and the other way round, in the journal’s style.
• Word count, abstract length, keywords and heading levels follow the author guidelines of the journal you chose.
• Declarations are ready: ethics approval, funding, conflicts of interest, data availability, author contributions.

Run a similarity check before submission and look at what it flags. Text reused from your own thesis or an earlier conference paper still counts; most journals expect you to rewrite it or disclose the earlier version. Our guide to why papers get desk rejected covers the technical faults that send papers back without review, and the cover letter guide covers the letter that goes with the submission.

Papers that do not follow IMRaD

Review articles are organised by theme, with a methods section describing the search if they are systematic. Case studies and case reports follow the case narrative. Humanities papers are usually argued essays with headings that follow the argument. The principle is the same: state the problem, show your evidence, say what it means.

If you want a subject specialist to review the structure and argument of your draft before you submit it, that is what our research paper writing support covers. The paper and its findings remain yours.