Most students start a research paper at the wrong end — opening a blank document and trying to write a first sentence before they know what they argue. That is why the paper stalls. A research paper is not a long essay; it is a structured argument backed by verified sources, and the people who finish on time start with the structure, not the sentence. This guide walks through the process from the first topic search to the final revision pass, with word-count targets you can use as checkpoints and a short example of what integrated sources look like in a paragraph. You will learn how to narrow a topic, build a working thesis, outline at the section level, draft the body first, cite sources honestly, and revise in passes that catch different problems.
PaperDraft is a writing assistant, not a paper generator — the draft is your starting point, not your submission. You are responsible for editing, verifying sources, and following your school's academic integrity policy.
What a research paper actually is
A research paper is a written argument that answers a focused question using evidence from verified sources, organized into a conventional structure (introduction, body, conclusion) and cited in a recognized style. The typical undergraduate research paper runs 1,800 to 3,500 words and draws on 8 to 15 sources. The graduate version is longer, cites more, and usually addresses a narrower question in more depth.
The genre is sometimes confused with adjacent types, and the distinctions matter for how you approach yours. A research paper builds an argument using existing sources — you are not reporting original experimental data. A research report or empirical paper presents original data you collected. A literature review synthesizes what has been said on a topic without advancing a new argument. An argumentative essay makes a case but may lean more on reasoning than sources.
A research paper sits in the middle: an argument, but one grounded in published evidence and positioned against what other scholars have said. The "research" part is not ornamental — it means your thesis has to engage with the literature, not sit on top of it.
Before you start
Most of the damage done to a research paper happens before the first sentence is typed. Three things need to be locked in before you draft.
Read the assignment brief like a contract
Your instructor's brief is the scoring rubric in disguise. Read it twice. Pull out the constraints — word count, citation style, number of sources required, whether primary sources are mandatory, whether a specific question is assigned or you choose one, the due date. If any of these are ambiguous, ask now, not the night before.
Narrow scope aggressively
The single biggest first-draft failure is a topic that is too broad. "Climate change" is not a topic; it is a field. You need a question you can answer in the word count assigned. A useful test: write your working question in one sentence. If you cannot state it in one sentence, it is still too broad. A 2,500-word paper needs a question that can be answered in 2,500 words — usually a specific policy, a specific event, a specific mechanism, a specific debate.
Confirm you can access the sources
Before you commit to a thesis, check that the literature is actually reachable. Search Google Scholar, your library catalog, and one subject-specific database (JSTOR for humanities, PubMed for biomedical, PsycINFO for psychology). If you cannot find at least 10 credible sources in 30 minutes of searching, the topic is either too narrow, too new, or in a field your library does not cover. Switch topics now rather than two days before the deadline.
Step-by-step: how to write a research paper
The following sequence has kept more papers on schedule than any single writing trick. Follow it in order.
1. Narrow your topic into a researchable question
Move from field to problem to question. A researchable question has a knowable answer, a bounded scope, and engages a real debate. "What causes inflation?" is a question but not researchable at paper length. "Did the 2021 US stimulus measurably contribute to core inflation in 2022?" is answerable with sources.
Common mistake: treating the topic as the question. "My topic is social media" is not a question. Push further until you can write the question as a sentence ending in a question mark.
Micro-example: Topic: youth mental health → Problem: social media's role in adolescent anxiety → Question: "Does time on Instagram correlate with anxiety symptoms in 14- to 17-year-olds in longitudinal studies?"
2. Do a preliminary literature sweep
Before you commit to an argument, read broadly. Skim abstracts and introductions of 8 to 15 sources. You are building a map, not memorizing content. Track what each source argues using a simple matrix: author, year, claim, method, relevance to your question. This is where you find the gap or tension your paper will address.
Common mistake: starting to write before reading. First drafts written without a literature sweep read like opinion pieces that happened to pick up citations on the way.
3. Write a working thesis statement
One sentence. It states what you argue, and ideally gestures at why it matters or how you will support it. "The 2018 EU ETS reform produced a measurable emissions reduction in the industrial sector, primarily through the market stability reserve mechanism, though effects on power generation were smaller than policy advocates predicted." That is specific, debatable, and tells the reader what will follow. See our thesis statement guide for the full treatment.
Common mistake: writing a topic sentence and calling it a thesis. "This paper will discuss the EU ETS" states the topic but takes no position.
4. Build a sectioned outline with word budgets
Take your thesis and break the argument into sections. A 2,500-word research paper might look like: intro (250w), background (400w), body section 1 (500w), body section 2 (500w), body section 3 (400w), counterargument (250w), conclusion (200w). Assign section word counts before drafting. If a section balloons past its budget, either cut or promote it to its own section. Our research paper outline guide has templates for common paper lengths.
Common mistake: outlining at the paragraph level before you have the section-level shape. Zoom out first.
5. Draft the body before the introduction
The introduction frames an argument you have not made yet, so drafting it first guarantees rewriting. Start with the body section where your evidence is strongest. Write through the middle sections in order, then the counterargument, then the conclusion. Write the introduction last, once you know what paper it is introducing.
Common mistake: spending two hours on an opening paragraph before you know what you argue. Write a placeholder intro ("In this paper I argue X") and move on.
6. Integrate sources with signal phrases and citations
Every claim that is not your own original argument and not common knowledge needs a citation. Use signal phrases to orient the reader: "Smith (2020) argues...", "Jones and Lee (2019) found that...", "Critics of this view, including Nguyen (2021), counter...". Paraphrase more than you quote; reserve direct quotes for language that cannot be rephrased without losing meaning. Format citations in the style your assignment requires — APA for psychology and most social sciences, MLA for humanities, Chicago for history.
Common mistake: quote-dumping without signal phrases. The reader cannot tell who is talking.
7. Revise in passes — argument, paragraphs, sentences
Do not mix levels of editing. First pass: does each section advance the thesis? Cut anything that does not. Second pass: does each paragraph have a clear topic sentence and a transition? Third pass: sentence-level clarity and citation formatting. Three passes catch more than one long pass.
Common mistake: fixing commas before the argument is solid. You will delete the sentence.
Stuck at the start? PaperDraft scaffolds a research paper draft — thesis, outline, opening sections — for you to revise. Start this paper — free.
Structure/outline template
For a 2,500-word undergraduate research paper, the following section structure gives you proportions that work. Adjust the counts up or down proportionally for shorter or longer papers.
Introduction (200–300 words). Hook, context, the question, the thesis in the last sentence, and a preview of the structure. The reader should leave the introduction knowing exactly what you argue and how you will proceed.
Background / literature context (350–450 words). The minimum a reader needs to understand your argument — key definitions, prior findings, the debate your paper joins. Do not write a literature review here; write the briefing a reader needs.
Body section 1 (450–550 words). Your first main point, with evidence, citations, and analysis. Structure each paragraph: topic sentence, evidence, analysis, link to the thesis.
Body section 2 (450–550 words). Second main point, same structure. The transition from section 1 should make the logical progression clear.
Body section 3 (350–450 words). Third main point or synthesis. Often this is where your own argument is sharpest because you have established the groundwork.
Counterargument (200–300 words). Name the strongest objection, acknowledge what it gets right, and explain why your thesis still holds. Papers without counterargument read as one-sided to graders.
Conclusion (150–250 words). Restate the thesis in light of the evidence, name the broader significance, and name an open question or implication. Do not summarize; synthesize.
References (separate page). Every in-text citation, none missing, none added. Use the correct hanging indent and style format.
Example excerpt
Here is what a single body paragraph looks like when the steps above are followed. The paragraph is from a paper on the EU ETS 2018 reform.
Early evidence suggests the 2018 reform's market stability reserve did bite where earlier phases of the ETS did not. Bayer and Aklin (2020) estimate that the reserve's removal of surplus allowances lifted the effective carbon price from around €8 per tonne in 2017 to above €25 by the end of 2018 — a shift they attribute specifically to expectations about the reserve rather than to broader macroeconomic conditions. Their difference-in-differences design compares ETS-covered installations to non-covered European manufacturing and finds industrial emissions falling at a 4.2% steeper annual rate in the covered sector after the reform (p. 8808). Two limitations are worth noting. First, the observation window is short — only the first full year post-reform. Second, concurrent policy changes at the member-state level, especially Germany's coal phase-out commitments, may confound the attribution. Still, the magnitude and timing make the reserve a plausible — and probably the primary — driver of the post-2018 price recovery.
Annotations: the paragraph opens with a claim (topic sentence), introduces the source with a signal phrase, gives a specific estimate with attribution, names the method, engages with limitations honestly, and ends with a measured synthesis. No quote dumps, no uncited assertions, no hedging to the point of meaninglessness.
Common mistakes
- No thesis, or a thesis that is actually a topic. "This paper discusses X" tells the reader nothing. Rewrite until the sentence states a position a reasonable person could disagree with.
- Body paragraphs without topic sentences. If a reader cannot tell what the paragraph is doing from the first sentence, the paragraph is doing too many things. Split it.
- Quote-dumping. Long block quotes with no analysis following them signal that the writer did not process the source. Paraphrase, then quote only what cannot be paraphrased.
- Unverified or hallucinated citations. Every citation must match a real source you actually read. Fabricated citations are an academic integrity violation at every institution; see our academic responsibility guide for the framing. This is the single most serious error in this list.
- Revising only at the sentence level. Commas cannot save a paper whose argument does not work. Revise from the outside in: argument, then paragraphs, then sentences.
How PaperDraft helps you start
PaperDraft is a writing assistant, not a paper generator — the draft is your starting point, not your submission. What it does well is get you past the blank page: it drafts a starting version of your thesis, builds a section outline with word-count ranges, and stubs opening paragraphs in academic register with citation placeholders in your chosen style. You then read, revise, verify every citation against a real source, and rewrite until the paper is yours. See our research paper landing for the full drafting flow, and research paper structure for a deeper look at section conventions.
Frequently asked questions
How many sources does a research paper need?
For an undergraduate paper of 2,000 to 3,500 words, 8 to 15 cited sources is typical. Graduate papers often cite 20 or more. Your assignment brief is the authority here — if it specifies a minimum, meet it; if it does not, aim for the density that lets each major claim rest on at least one cited source.
How long should a research paper be?
Undergraduate research papers usually run 1,800 to 3,500 words. Term papers in upper-division courses can reach 5,000. Graduate seminar papers commonly hit 6,000 to 10,000. The assignment brief sets the target; treat the word count as a budget, not a minimum to pad.
What citation style should I use?
The style depends on the discipline. Psychology, education, and most social sciences use APA. Humanities typically use MLA. History and some humanities subfields use Chicago. Engineering uses IEEE. If the brief does not specify, ask your instructor — do not guess.
Can I use AI to help with my research paper?
Many institutions now allow AI-assisted drafting if disclosed. Policies vary widely — some departments prohibit it for certain assignments, others encourage it as long as you disclose use and verify all content. See our AI disclosure guide for how to handle this honestly, and always check your specific course and institutional policy before using any AI tool.
How do I avoid plagiarism in a research paper?
Cite every claim that is not your own original reasoning and not common knowledge. Paraphrase in your own words and still cite. Quote directly only what cannot be rephrased, and always with a citation. Our avoid plagiarism guide walks through the specific habits that keep you in the clear.
When should I write the introduction?
Last. Write the body sections first, then the conclusion, then the introduction. You can only introduce a paper you have already written. A placeholder opening ("In this paper I argue...") is fine as a scaffold while you draft the middle.