The conclusion is the shortest chapter in most dissertations and the one markers read most carefully, because it is where you stop reporting and start claiming. Everything before it — literature, methodology, findings — builds the evidence; the conclusion is where you tell the reader what that evidence adds up to: whether your research question is answered, what your study contributes, and where its limits lie. Students underwrite it because they are exhausted, or overwrite it by re-arguing the whole dissertation. Both mistakes cost the same marks. This guide gives you the structure that works, section by section, with the pitfalls marked. It completes the chapter-by-chapter set that starts with our dissertation structure guide — and if the methodology chapter is still causing trouble, that guide’s companion on how to write a dissertation methodology covers it in the same detail.
How long should a dissertation conclusion be?
Typically 5–10% of your total word count — around 500 to 1,000 words in a 10,000–15,000-word dissertation, sometimes a little more where recommendations carry real weight (business and professional doctorates’ applied projects tend to run longer). If yours is ballooning past that, you are probably re-writing the discussion; if it fits on one page, you are almost certainly under-claiming. Set the discussion and conclusion boundary deliberately: the discussion interprets each finding against the literature; the conclusion steps back and states the whole. Some universities merge them into one chapter — if yours does, the functions below still all need doing, in the same order.
The structure that works
1. Restate the aim and answer the research question — directly
Open by returning to your research question or aim, then answer it in plain terms. Not “this dissertation set out to explore…” followed by fog: state what you found, at the level of your headline results. A marker should be able to read your first paragraph and know what your study concluded. If you had objectives or sub-questions, answer each briefly — a sentence or two per objective, mapped to the finding that settles it. This mapping is the most commonly skipped element and the one examiners check first, because an unanswered objective is unfinished business hanging over the whole grade.
2. Summarise the key findings — the distilled version
Two to four findings, the ones that matter, stated without the data. No tables, no statistics beyond the single most important figure, no new quotations. You are not repeating the results chapter; you are declaring what survives of it. If a finding does not connect to your research question or your contribution, it does not belong here, however hard you worked on it.
3. State your contribution
What exists now — in knowledge, in practice, in method — that did not exist before your project? Contributions come in kinds: confirming or challenging an established theory in a new context; evidence for a setting the literature neglected (a sector, a company size, a patient group); a practical output (a framework, a set of recommendations a real organisation could use); occasionally a methodological point. Undergraduate and Masters contributions are modest and markers know it — “this study provides the first evidence on X within Y setting” is a genuine contribution; claiming to have transformed the field is not. Precision about a small contribution beats inflation of a large one, every time.
4. Acknowledge limitations — honestly, briefly, without self-sabotage
Sample size and who it excluded, single-site or single-organisation scope, self-reported data, cross-sectional timing, researcher position in insider studies. Name each real limitation in a sentence, note its direction of effect (“this means the findings may understate…”), and move on. Two disciplines matter here: never introduce a limitation so severe it retrospectively invalidates your findings (if it genuinely does, that belongs in the discussion, argued properly), and never list ritual limitations (“time was limited”) that carry no analytical consequence.
5. Recommendations and future research
Two kinds, kept separate. Recommendations for practice or policy flow from your findings — each one should trace to a result, be addressed to someone who could act on it, and be feasible. Recommendations for future research flow from your limitations: the larger sample, the longitudinal design, the comparison setting your study points to but could not reach. Examiners read future-research sections as a test of whether you understand your own design; generic “more research is needed” fails that test, specific proposals pass it.
6. A closing statement that lands
One short final paragraph: the significance of the work, stated once, in your strongest sentence. No new material, no quotation you have not used before, no apology. Some students end by returning to the problem that motivated the study — the gap in the literature, the issue in practice — and stating what your work now says about it. That circular shape, problem to answer, is why conclusions written this way feel finished.
Conclusions by dissertation type
The structure above flexes by project type. Quantitative studies answer the question through their hypotheses: state plainly which were supported, which were not, and what the pattern means — examiners dislike conclusions vaguer than the statistics beneath them. Qualitative studies conclude in themes and meaning: what the participants’ experience, taken together, reveals about the phenomenon, and how that extends or complicates the existing account. Business and consultancy projects weight recommendations far more heavily — sometimes a third of the chapter — because the client organisation is a real reader; feasibility, cost and sequence belong here, not as afterthoughts. Systematic reviews conclude by synthesising across studies: the direction and consistency of the evidence, the quality caveats that condition it, and what practitioners or researchers should do differently on Monday morning. Mixed-methods projects owe the examiner one extra move: stating what the two strands concluded together that neither concluded alone — the integration is frequently the project’s real contribution, and the conclusion is its only natural home.
Sentences that quietly weaken conclusions
Markers meet the same self-inflicted wounds every season. “This dissertation has been a valuable learning experience” — your learning is assessed elsewhere; the conclusion is about the research. “Due to time constraints, the study could not…” — constraints explain a design, they do not excuse one, and examiners read the phrase as an apology in advance. “The findings prove that…” — empirical studies support, suggest and indicate; they almost never prove, and overstating certainty invites the examiner to test it. “More research is needed” standing alone — needed by whom, into what, using which method? And the conclusion that ends on a limitation, leaving the reader’s final impression as a list of flaws: always close on contribution or significance, with limitations handled earlier in the chapter. None of these sentences is fatal alone; together they signal a writer who stopped trusting their own work at exactly the moment the work needed presenting.
What never goes in a conclusion
New evidence, new literature, new analysis — the conclusion’s most serious sin, because unexamined material cannot be credited and signals you do not know where your argument ends. Direct quotations from participants or sources (their place was the findings). Apologies and confessions (“despite the small sample, I tried my best”). And a summary of every chapter — the “in chapter one I…, in chapter two I…” paragraph belongs to school projects, not dissertations.
A worked example of the opening move
Weak: “This dissertation has explored the topic of hybrid working and job satisfaction. Many interesting findings were discovered and discussed.” Strong in shape: “This study set out to establish whether hybrid working patterns affect job satisfaction among UK professional-services staff. The evidence supports a conditional answer: satisfaction was significantly higher under hybrid arrangements, but only where employees controlled their own schedules — imposed hybrid patterns showed no benefit.” The second version answers, qualifies, and tells the marker exactly where the argument is going. Write your opening paragraph to that standard and the rest of the chapter has a spine.
If you are writing the conclusion before the rest feels finished
That instinct is sound — conclusions are best drafted once findings exist but revised last of all, after the discussion has settled. Write the answer to your research question as it currently stands, draft the contribution you believe you have, and return at the end to check every claim in the conclusion is earned by a chapter beneath it. That claim-by-claim audit — conclusion statements on one side, the findings that support them on the other — is the single highest-value hour you can spend in your final week.
When the chapter will not come together
A conclusion resists writing when the findings themselves are unsettled — the analysis is not finished, or the discussion never reached positions the conclusion could summarise. If that is your situation, our dissertation chapter help service exists for exactly this stage: send the chapters you have and we will show you, in a model chapter, what yours needs to claim. And the dissertation help page covers support for the project as a whole, every chapter referenced precisely and written from scratch, from £20 per 1,000 words.
Dissertation conclusion questions
Should I mention a finding that contradicts my original hypothesis?
Yes, and you should deal with it directly rather than quietly leaving it out. Stating plainly that the evidence did not support part of your expectation, and explaining what that tells you about the topic, is treated as a sign of a careful researcher. Markers are more concerned by a conclusion that claims complete success than by one that reports a mixed picture honestly.
Can I repeat a limitation that I already described in the methodology?
You can return to it briefly, but the job changes in the conclusion. In the methodology you explain what you decided and why; here you explain what effect that limitation had on how far your findings can be generalised. Keep it to a sentence or two and pair it with what you would do differently next time.
How specific should my suggestions for future research be?
Fully specific beats vague every time. Instead of writing that further research is needed, name the population, setting or variable you would change and why that change would answer the gap your study exposed. A precise suggestion proves you understand the limits of your own design.
Is it a good idea to finish the conclusion with a quotation?
Usually not. A borrowed quotation can make the final line sound decorative rather than decisive, and it shifts the last voice the marker hears away from your own judgement. End instead with your own closing statement that returns to the research question and states, in one clear sentence, what your project has shown.