How to Write a Conclusion for a Research Paper

The conclusion is the final section of a research paper, where you state what your research has established and why it matters. If the introduction asks the question and the body supplies the evidence, the conclusion delivers the answer and steps back to show its significance.

A good conclusion is not a summary of everything in the paper. Readers have just finished the body; they do not need it repeated. What they need is synthesis: the main points brought together into a clear answer, plus a sense of what that answer changes for the field, for practice or for future research.

In empirical papers following the IMRaD format, the conclusion is often short, sometimes a single paragraph at the end of the discussion, and some journals do not require a separate conclusion at all. In argumentative papers, essays and term papers, the conclusion is always its own section, typically the final paragraph or two. Our guide to writing a research paper shows where it fits in the overall structure.

Most strong research paper conclusions contain five elements, though short papers may combine some of them into a single sentence.

In empirical papers where the discussion has already covered limitations in detail, the conclusion can mention them in a single clause or omit them. In humanities papers, the closing element is more often a broader implication than a call for future research.

Knowing what not to write is as important as knowing what to include. Leave out:

• New evidence or arguments. If a point is important enough to support your answer, it belongs in the body. The conclusion draws on what you have already shown.
• New citations, as a rule. Referring back to a key source already discussed is fine, but introducing new literature signals that the discussion was incomplete.
• Verbatim repetition. Copying your thesis or abstract word for word reads as filler.
• Detailed statistics. Restate findings in words ("students who received weekly feedback were more likely to complete the course"), not with test statistics and p values.
• Overclaiming. Do not suddenly claim that your study "proves" something or "solves" a problem when the body made more careful claims.
• Apologies and hedging spirals. "This is only one perspective, and I may be wrong" undermines your work. Limitations should be specific and measured, not defensive.
• Sentimental or dramatic endings. Ending with a quotation, a rhetorical question or a call to save the world rarely suits academic writing.

If you find conclusions hard to write, it is usually because you are trying to write them from the draft rather than from your argument. This short process works for most papers:

1. Reread your introduction. Copy out your research question or thesis and the gap you identified. The conclusion must answer exactly this question, not a slightly different one that the paper drifted toward.
2. Write the answer in one sentence. Without looking at the draft, state what you found or argued. If you cannot, revisit the discussion before writing the conclusion.
3. List your three most important findings or arguments. Then ask how they relate: do they reinforce each other, qualify each other or add up to something none shows alone? That relationship is your synthesis.
4. Ask "so what?" twice. The first answer is usually obvious ("the results are relevant to managers"). The second answer is usually your real significance ("team schedules should be designed with newcomers' learning in mind, not only individual preference").
5. Choose one specific next step. A precise future question ("whether the effect holds for part-time staff") is more useful than "more research is needed."
6. Draft, then cut. Remove any sentence that repeats the body without adding synthesis, and any sentence that introduces something new.

Conclusion ideas often come from the limitations and surprises you noticed while analyzing results. A finding that did not fit your expectations can become the most interesting closing point, provided you have already discussed it in the body.

A reliable structure for a conclusion paragraph moves in the opposite direction from the introduction: from specific back to general. The introduction narrows from context to your question; the conclusion starts with your answer and widens to its implications.

1. Opening sentence: answer the research question or restate the thesis in new words.
2. Two to four sentences: synthesize the main findings or arguments, showing how they fit together.
3. One to two sentences: explain the significance.
4. One sentence (optional): note the main limitation.
5. Closing sentence: a specific direction for future research or a broader implication.

Conclusion template for a research paper

Use this as scaffolding for a first draft, then rewrite it in your own voice so it does not read as formulaic:

 This [study / paper] set out to [aim]. The [results / analysis] show that [main answer]. In particular, [key finding 1] and [key finding 2], which together suggest [synthesis]. These findings matter because [significance for theory, practice or policy]. Although [main limitation], the [study / argument] indicates that [broader point]. Future research should [specific next question]. 

For a longer paper or a thesis chapter, each step can become its own paragraph, and the significance and future research steps often expand the most.

Restating a thesis means returning to your central claim with the benefit of everything you have shown. The restated version should be recognizably the same claim, but more confident and more precise, because it is now supported by evidence.

The strong versions do three things: they use new wording, they fold in the evidence, and they add nuance that the original could not have contained before the research was done.

A transition sentence into the conclusion signals that the argument is complete and that you are now stepping back. In many papers the section heading does this work, so an explicit transition is optional. When you do need one, choose something that links to your content rather than a stock phrase.

• "Taken together, these findings suggest..."
• "The evidence considered here points to..."
• "Returning to the question that opened this paper..."
• "What emerges from this analysis is..."
• "Overall, the results indicate..."

"In conclusion" and "to sum up" are not wrong, but they are overused and redundant under a heading labeled Conclusion. Good concluding statements for essays and papers avoid announcing that the end has arrived and instead deliver the final point directly. The last sentence should be something only your paper could end with: a specific implication, a precise open question or a sharpened version of your claim.

The three examples below are original and written for this guide. Each is annotated with the elements from the table above.

Example 1: Empirical social science paper

[Answer] This study asked how first-year employees in hybrid teams acquire the informal knowledge needed to do their jobs. [Synthesis] Interviews showed that new hires learned most when teams coordinated at least one shared office day and when managers made informal questions explicitly welcome; where schedules were left entirely to individual choice, new employees reported relying on trial and error. [Significance] These findings suggest that the flexibility valued by experienced staff can come at a hidden cost to newcomers, and that team-level scheduling decisions are also learning decisions. [Limitation] Because the study drew on four organizations in one sector, the patterns may differ elsewhere. [Future direction] Longitudinal research following new hires across their first year could test whether coordinated schedules lead to faster independence on the job.

Example 2: Lab or field science paper

[Answer] Bee species richness declined with green roof height, and the decline was steeper for small-bodied species. [Significance] For cities aiming to support pollinators, this implies that low roofs near ground-level green space are likely to deliver more benefit than tall roofs planted with the same species. [Future direction] Whether connecting tall roofs with intermediate-height plantings can offset the height effect remains an open question for urban planners and ecologists.

This one is deliberately short. In IMRaD papers, the discussion has already interpreted results and covered limitations, so the conclusion only needs to deliver the answer and its practical consequence.

Example 3: Humanities argumentative paper

[Restated thesis] Read as performances of respectability rather than as transparent reports, emigrant letters tell a different story about migration. [Synthesis] The family correspondence examined here shows writers adjusting their accounts of prosperity as their readers changed, inflating success for parents and village neighbors, and admitting hardship only to siblings who had also emigrated. [Significance] Historians who use such letters as evidence of economic motives should therefore treat claims about wages and land as statements shaped by audience. [Closing thought] The letters remain invaluable sources, but what they record most reliably is not how emigrants lived; it is how they wanted to be seen.

In empirical papers, the discussion and conclusion are closely related, and some journals combine them. They do different jobs, though, and confusing them is one of the most common structural problems in student papers.

A useful test: if a paragraph compares your result with a specific earlier study, it belongs in the discussion. If it tells the reader what the whole study means, it belongs in the conclusion. When your journal or instructor asks for a combined "Discussion and Conclusion" section, keep the same logic and put the concluding material in the final paragraph.

The same principles apply to other academic genres, with a few adjustments:

• Essays and term papers: the conclusion is usually one paragraph. Restate the thesis in new words, connect the main arguments and end with an implication. A college-level conclusion paragraph goes beyond summary to explain why the argument matters.
• Literature reviews: to conclude a literature review, summarize the main patterns across the literature, state the most important gaps or unresolved debates and, if the review precedes your own study, show how your research addresses those gaps. See literature review outline for structure.
• Theses and dissertations: the conclusion is a full chapter. It typically answers each research question in turn, states the contribution to knowledge, discusses implications and limitations and recommends specific future research.
• Research reports: conclusions are often followed by numbered recommendations, each tied to a finding.

A conclusion is usually 5 to 10 percent of the paper. For a 3,000-word paper that is roughly 150 to 300 words, often a single paragraph. Journal articles typically have a conclusion of one to three paragraphs, and a thesis has a full chapter. If your conclusion is longer than your introduction, it is probably repeating the discussion.

Before you submit, check that your conclusion:

1. Answers the research question or restates the thesis in new words.
2. Synthesizes rather than lists the main findings.
3. Explains why the answer matters.
4. Introduces no new evidence, arguments or sources.
5. Matches the strength of claims made in the body.
6. Ends with a specific final point, not a generic one.

If you would like feedback on whether your conclusion does its job, our academic editing service reviews structure and clarity while you remain the author.