How to Write a Research Paper (2026 Complete Guide)

Writing a Research Paper: What You Actually Need to Know

Writing a research paper is one of the most valuable skills in academic and professional life. It is also one of the most misunderstood. Most guides focus on formatting rules or citation styles — useful, but not the core of what makes a paper work.

This guide covers the entire process from scratch: how to choose a topic with the right scope, how to review existing literature efficiently, how to build an argument, how to structure each section, how to write clearly under academic conventions, and how to get through revision and submission.

Whether you are writing your first undergraduate paper, a graduate thesis chapter, or a manuscript for journal submission, the process is largely the same. The scale differs; the logic does not.

This is a practical guide — not a style manual. It focuses on the decisions that actually determine whether a paper succeeds.

The Research Paper Process at a Glance

Before diving in, it helps to understand the full arc of the process. Writing a research paper is not linear — you will loop back through earlier stages repeatedly. But the stages have a natural sequence.

Step 1 — Choose a Topic with the Right Scope

The most common mistake in early research papers is choosing a topic that is too broad. "Climate change and health" is not a research paper topic — it is an entire field. "The effect of extreme heat events on emergency department visits in urban areas: a systematic review of studies from 2010–2024" is a topic.

What makes a good research topic?

A good research topic sits at the intersection of three things: something you find genuinely interesting, something that has existing literature to engage with, and something with a clearly articulable gap or question that your paper will address.

How to narrow a broad topic

Start with your broad area of interest. Then ask: what specific aspect, population, time period, intervention, or question interests you most? Apply each narrowing dimension until the topic becomes a specific, answerable question.

Example narrowing: - Broad: "Social media and mental health" - Narrowed by population: "Social media use and mental health in adolescents" - Narrowed by outcome: "Social media use and depression symptoms in adolescents" - Narrowed by study type: "Does reducing social media use reduce depression symptoms in adolescents? A review of experimental studies"

This last version is a research paper topic.

Check what exists first

Before committing to a topic, do a preliminary literature search to confirm: (1) enough prior work exists to build on, and (2) your specific angle has not already been exhaustively covered. Use Ocean of Papers to search across PubMed, arXiv, and OpenAlex simultaneously — this gives you the broadest view of what exists in your area before investing time in a topic that is either too crowded or too thin.

Step 2 — Conduct Your Literature Review

The literature review has two functions: it shows you what is already known (so you do not reinvent the wheel), and it establishes the context your paper will respond to (so readers understand why your contribution matters).

How much literature do you need?

This depends on the paper type and level. For an undergraduate paper, 10–20 sources may be sufficient. For a graduate thesis chapter or journal article, 30–80 is more typical. For a systematic review, hundreds may be relevant. The goal is comprehensiveness within scope — not exhaustiveness across all of human knowledge.

A systematic approach to literature search

Do not just search once and stop. Good literature review requires multiple rounds:

1. Seed search — Use 2–3 core keyword combinations to find the 10–15 most relevant papers
2. Snowball — Read those papers' reference lists and find additional relevant sources
3. Forward citation — Use tools like Ocean of Papers' related papers feature or Semantic Scholar to find newer papers that cite your core sources
4. Gap check — Search specifically for the gap your paper addresses, to confirm it is genuinely underexplored

Organizing what you find

As you read, take notes structured around: (1) the main argument or finding, (2) the methodology, (3) the limitations, and (4) how it relates to your question. A simple spreadsheet works. Reference management tools like Zotero or Mendeley can store PDFs alongside your notes and auto-generate citations.

Synthesizing, not summarizing

The literature review section of your paper is not a list of summaries ("Smith (2019) found that X. Jones (2021) found that Y."). It is a synthesis — grouping sources by theme, showing where they agree, where they disagree, and where the gaps are. The literature review should build to your research question as its logical conclusion.

1. Build your keyword list — Generate synonyms, related terms, MeSH terms (for medical topics), and Boolean combinations
2. Search across multiple databases — Use Ocean of Papers to cover PubMed, arXiv, OpenAlex, and Europe PMC in one search
3. Download and organize PDFs — Save to Zotero or Mendeley with citation metadata — do not store loose PDFs
4. Read strategically — Abstract first, then conclusions, then methods — only full-read papers directly relevant to your question
5. Code and cluster findings — Group papers by theme, methodology, finding direction, or population
6. Map the gaps — Identify what questions remain unanswered — your paper should address at least one

Step 3 — Define a Sharp Research Question

Your research question is the single most important sentence in your paper. Everything else — your literature review, methodology, analysis, and conclusion — exists to address it.

What makes a good research question?

A strong research question is: - Specific — It asks about a defined phenomenon, population, intervention, or relationship - Answerable — It can be addressed with evidence, analysis, or argument within the scope of the paper - Significant — The answer matters — it advances understanding or has practical implications - Novel — It has not already been definitively answered in the existing literature

The PICO framework (for empirical/medical research)

For clinical, public health, or social science papers with an intervention focus, the PICO framework structures a strong research question: Population, Intervention, Comparison, Outcome. Example: In adults with type 2 diabetes (P), does dietary intervention (I) compared to standard care (C) reduce HbA1c levels at 6 months (O)?

For theoretical or humanities papers

The central question is often about interpretation, causation, relationship, or argument. Example: "How did Cold War political pressures shape the development of US federal science funding policy between 1945 and 1960?"

Thesis statement vs. research question

In empirical papers, the research question is usually stated as a question and answered in the conclusion. In argumentative papers (common in humanities and some social sciences), the research question is often reformulated as a thesis statement — a declarative claim your paper will defend. State it clearly in the introduction.

Step 4 — Structure Your Paper

Different paper types have different standard structures. Using the expected structure for your paper type makes it easier for readers to navigate and for reviewers to evaluate.

The IMRaD structure (empirical papers)

IMRaD — Introduction, Methods, Results, and Discussion — is the standard structure for empirical research papers in science, medicine, psychology, and social science. It mirrors the scientific process:

• Introduction — Background, existing literature, gap, research question/hypothesis
• Methods — How the study was conducted (design, participants, data collection, analysis) — in enough detail for replication
• Results — What was found (facts only, no interpretation)
• Discussion — What the results mean, limitations, comparison to existing literature, implications
• Conclusion — Summary answer to the research question, future directions

For literature review papers

Introduction → Methods (search strategy and inclusion criteria) → Results (synthesis of findings by theme) → Discussion → Conclusion

For argumentative/theoretical papers

Introduction (with thesis) → Background/context → Main argument (2–4 body sections) → Counter-arguments and responses → Conclusion

Always include: - Abstract (100–300 words summarizing the whole paper) - References (formatted in the required style) - Appropriate section headings

Step 5 — How to Write Each Section

The Abstract Write the abstract last. It should answer: what is the problem, what did you do, what did you find, and why does it matter — in 150–300 words. Every word counts. Avoid jargon that is not explained elsewhere. The abstract is the most-read part of any paper.

The Introduction Open with the broad context (1–2 sentences). Narrow to the specific problem. Review the key existing literature. Identify the gap. State the research question and, if applicable, the hypothesis or thesis. End with a brief signpost of the paper's structure. The introduction is a funnel: broad to narrow.

The Methods Write with enough precision that another researcher could replicate your study. Include: study design, participants or data sources, data collection procedures, variables, and analysis methods. Use past tense. For literature reviews, describe your search strategy (databases, keywords, date range, inclusion/exclusion criteria).

The Results Present findings without interpretation. Use past tense. Refer to figures and tables in order (Figure 1 before Figure 2). Report statistics with appropriate precision and include confidence intervals where relevant. Do not interpret — save that for the Discussion.

The Discussion This is where the paper lives or dies. Start by directly answering your research question. Then: explain what your findings mean, compare them to existing literature (agree, disagree, extend), acknowledge limitations honestly, discuss implications, and suggest future research. Avoid overstating conclusions — the data supports what it supports.

The Conclusion 1–2 paragraphs. Restate the research question and give the answer. Summarize the key contribution. Note the most important implications. Do not introduce new information.

References Use a reference manager. Cite as you write — do not reconstruct references at the end. Follow the required style guide (APA, MLA, Chicago, Vancouver, Harvard, etc.) precisely.

Step 6 — Write Clearly and Precisely

Academic writing has a reputation for being dense and difficult. The best academic writing is actually very clear — it just deals with complex ideas. Clear writing is not simpler thinking; it is more precise expression of the same thinking.

Principles of clear academic writing:

One idea per paragraph. Each paragraph should have a single main idea, stated in the topic sentence. Everything else in the paragraph should support that idea. If you find yourself covering two different ideas in one paragraph, split it.

Use active voice where possible. "We found that temperature increased" is clearer and more honest than "It was found that temperature increases were observed." Passive voice is appropriate when the actor is unknown or irrelevant, but do not hide agency in your own work.

Define technical terms. Do not assume readers know specialized terminology. Define it on first use. After defining, use the term consistently — do not vary terminology for stylistic reasons.

Avoid hedging stacks. One hedge is appropriate when uncertain: "This suggests that..." But "It would seem that this might possibly indicate that perhaps..." signals unconfident writing. Hedge once, clearly.

Vary sentence length. Long sentences carry complex ideas. Short ones land emphasis. Vary them. Reading a page of identically-structured medium-length sentences is exhausting.

Cut ruthlessly in revision. The first draft will be too long. Every word that does not do work should go. Academic word counts are limits, not targets.

Step 7 — Citations, References, and Academic Integrity

Correct citation is not just a formatting requirement — it is fundamental to academic integrity and to the scientific record. Citations do three things: they acknowledge the intellectual debt owed to prior work, they allow readers to trace and verify your sources, and they connect your paper to the larger scholarly conversation.

Cite as you write The most common citation error is reconstructing references at the end. Use a reference manager (Zotero, Mendeley, EndNote) from the start. Paste DOIs or drag PDFs in — the metadata is extracted automatically. Generate citations in the required style with one click.

What to cite Cite any specific claim, fact, statistic, figure, or idea that is not your own original contribution. Cite methods you are using but did not invent. Cite tools and software. When in doubt, cite.

Common style guides - APA (American Psychological Association) — psychology, social sciences, education - MLA (Modern Language Association) — humanities, literature, language arts - Chicago/Turabian — history, humanities, arts (two systems: notes-bibliography and author-date) - Vancouver — medicine, clinical sciences (numbered references) - Harvard — broad use in UK and Australia across many disciplines

Avoiding plagiarism Plagiarism — presenting others' work as your own — is a serious academic offense. This includes copying text without quotation and citation, paraphrasing without citation, and self-plagiarism (reusing your own prior work without disclosure). When in doubt, cite. When paraphrasing, change both the words and the sentence structure — do not just swap synonyms.

Ocean of Papers provides formatted citations (APA, BibTeX, RIS) directly on each result. Click the citation button on any paper to copy the reference.

Step 8 — Revise, Edit, and Proofread

First drafts are not papers — they are raw material. Revision is where papers are actually written.

Three-pass revision

Do not try to fix everything at once. Use three passes, each with a different focus:

Pass 1 — Structural revision Read the whole paper from start to finish. Does the argument flow logically? Does each section do what it should? Is the research question clearly answered? Does anything belong somewhere else? Delete or restructure at this level before fixing prose.

Pass 2 — Paragraph and sentence revision Work paragraph by paragraph. Does each paragraph have a clear topic sentence? Is every sentence doing work? Cut redundancy. Strengthen weak hedges or vague claims. Check that evidence supports claims.

Pass 3 — Proofreading Fix spelling, grammar, punctuation, and formatting. Check that all citations in the text appear in the reference list and vice versa. Verify page limits, heading formats, and any submission requirements.

Get feedback before final revision Ask someone else to read it — a peer, a supervisor, a writing center consultant. Fresh eyes catch things you have become blind to. Ask them specifically: Is the argument clear? Is anything confusing? Does the conclusion answer what the introduction promised?

Common revision mistakes - Only fixing grammar without addressing structural problems - Polishing the introduction before the argument is settled - Treating word-count limits as targets rather than maximums - Submitting without reading the final version aloud at least once

The Most Common Research Paper Mistakes (and How to Avoid Them)

1. Topic too broad The most frequent issue, especially at undergraduate level. If your topic covers an entire field, you cannot say anything meaningful in a 10-page paper. Narrow until you have a specific, answerable question.

2. Literature review as a list Stringing summaries together is not a literature review. Group sources by theme or argument. Show how they relate to each other and to your question. Build toward the gap your paper addresses.

3. Methods section too vague to replicate If someone else cannot reproduce your study from your methods section, it is not good enough. Every key decision — why this design, why this sample, why this analysis — should be explained.

4. Results and discussion mixed together Results present what you found. Discussion interprets what it means. Keep them separate unless your paper type explicitly merges them (some short papers use a combined Results and Discussion section).

5. Over-claiming in the conclusion Your data supports what it supports. Generalizing beyond your sample, overstating causal claims from correlational data, or ignoring limitations in your conclusion are all serious errors. Strong conclusions are precise, not dramatic.

6. Leaving citations to the end Reconstructing 40 references from scratch is slow, error-prone, and painful. Use a reference manager from day one.

7. First draft submitted as final Virtually no first draft is a final paper. Budget time for at least two rounds of serious revision.

Where to Find High-Quality Sources for Your Paper

The quality of your sources directly affects the credibility of your paper. Here is where to find peer-reviewed, citable literature:

Ocean of Papers (oceanofpapers.com) Searches PubMed, arXiv, OpenAlex, bioRxiv, medRxiv, and Europe PMC simultaneously. Automatically surfaces free PDF links for roughly half of all papers. The fastest starting point for multi-database literature search. No account required.

PubMed / PubMed Central The definitive database for biomedical and life sciences literature. PubMed indexes abstracts and metadata; PubMed Central includes full-text free-access articles.

arXiv (arxiv.org) Essential for computer science, mathematics, physics, and quantitative biology. Virtually all papers in these fields are available as free preprints the day they are submitted.

Google Scholar Broad coverage across all disciplines. Useful for finding older papers and grey literature. Less reliable for systematic review because coverage is not standardized.

Web of Science / Scopus Premium academic databases with citation tracking. Often available through university libraries. Useful for systematic searches requiring traceable methodology.

Evaluating source quality Prioritize peer-reviewed journal articles over preprints when possible. Check the journal's impact factor and reputation in the field. Be cautious with predatory journals — use Think Check Submit (thinkchecksubmit.org) to verify journals you are not familiar with.

Frequently Asked Questions

How long should a research paper be?

Length depends on the type of paper and the requirements. Undergraduate papers are typically 1,500–5,000 words. Graduate papers and journal articles commonly range from 4,000–10,000 words. Conference papers are usually 4,000–8,000 words. Systematic reviews can exceed 10,000 words. Always check the specific requirements before writing — length is a constraint, not a goal.

How do I start a research paper if I do not know where to begin?

Start with the literature, not the writing. Do a preliminary search on your broad topic using Ocean of Papers or Google Scholar. Read 5–10 recent papers on the topic. This will clarify what is already known, reveal the key debates, and surface a gap or question you can address. Once you can articulate the gap, the direction of the paper becomes clear.

What is the difference between a research paper and an essay?

An essay expresses and defends an argument based on analysis or interpretation. A research paper makes an original contribution — empirical, theoretical, or analytical — to a defined body of knowledge, grounded in systematic engagement with existing literature and evidence. Research papers typically follow established structural conventions (IMRaD or equivalent), include a formal methods section, and cite primary research sources.

How many sources should a research paper have?

There is no fixed number. Undergraduate papers typically have 10–25 sources. Graduate and journal-level papers commonly cite 30–80 sources. Systematic reviews may include hundreds. The right number is the minimum required to: (1) situate your work in the existing literature, (2) support all significant claims, and (3) demonstrate breadth of engagement with the field.

Can I use Wikipedia as a source?

Not as a primary citation. Wikipedia is useful for background understanding and for finding primary sources (check the reference section of Wikipedia articles), but it is not a citable academic source in most contexts. Cite the original research or authoritative sources that Wikipedia itself cites.

What is a peer-reviewed paper and why does it matter?

Peer review is a quality control process in which a submitted paper is independently evaluated by other experts in the field before publication. The reviewers assess the methodology, validity of conclusions, and significance of the contribution. Peer-reviewed papers are the gold standard of academic evidence — they have passed independent expert scrutiny. When searching for sources, prefer peer-reviewed journal articles over non-peer-reviewed reports, blog posts, or preprints where possible.

How do I avoid plagiarism?

Cite every claim, idea, statistic, or piece of evidence that is not your own original contribution. When paraphrasing, change both the words and the sentence structure — and still cite the source. Use quotation marks and a citation for any direct quotation. Use plagiarism detection tools (Turnitin, iThenticate) before submission to catch unintentional issues. When in doubt, cite.

How do I write a research paper abstract?

Write the abstract last. Include: (1) the problem or research question, (2) your approach or methods, (3) the key findings or argument, and (4) the significance or implications. Keep it within the word limit (usually 150–300 words). Every sentence should carry information — cut any sentence that says nothing specific. Do not include citations in the abstract.

What citation style should I use?

Use whatever style is required by your course, institution, or target journal. If not specified: APA is common in social sciences; MLA in humanities; Chicago in history; Vancouver in medicine; IEEE in engineering. Choose one and apply it consistently throughout.

How do I find a good research paper topic?

Start with a broad area you find genuinely interesting. Search that area in Ocean of Papers to understand what research exists. Read 5–10 recent papers to understand the current debates and open questions. Ask: what question do these papers leave unanswered? What aspect has not been studied with this population, methodology, or in this context? The intersection of your interest and a genuine gap is a good topic.

Getting Started: Your First Research Paper in Steps

Writing a research paper is a learnable skill. Every researcher — every professor who publishes regularly — once wrote a confusing first paper. The process improves with each paper you write.

Here is the practical starting point:

1. Search your broad topic in Ocean of Papers — read 5–10 papers to understand the landscape
2. Identify one specific question those papers do not fully answer
3. Articulate that question precisely
4. Read more deeply in that specific area — 20–30 sources
5. Plan your structure (IMRaD for empirical, argumentative for theoretical)
6. Write the methods or body sections first — the introduction last, abstract last
7. Revise in three passes: structure, then paragraphs, then proofreading
8. Get feedback, revise again, then submit

The literature is the foundation. Start there — and Ocean of Papers makes that part faster than it has ever been.

Related guides

• Research paper examples (annotated, free PDFs)
• How to download research papers for free
• Best Google Scholar alternatives
• Best literature mapping tools compared
• Swipe mode: build a reading list faster

Search free research papers · Build a citation map from any search result · Browse all research guides