How to Write a Dissertation Discussion Chapter: Turn Your Results Into an Argument

Your results chapter is finished, the tables look tidy, and the next heading in your contents list says Discussion. That is where many students stall, because the work shifts from reporting what happened to arguing what it means. Learning how to write a dissertation discussion chapter comes down to one habit: take each finding, explain it, test it against the literature and say what follows from it.

Settle these questions before you draft a word:

• Which of my findings answer a research question directly, and which are side results?
• Does each finding agree with, contradict or extend the studies in my literature review?
• What could explain the result I did not expect?
• Which limitations change how far my conclusions can reach?
• What can I claim outright, and what should I only suggest?

Most of the difficulty fades once you know the chapter’s job, where it stops, and a repeatable pattern for each paragraph. Conventions vary by discipline and institution, so use the structure below as a sound default and check your module handbook for anything specific.

What a Dissertation Discussion Chapter Is Meant to Do

A discussion chapter interprets your findings, judges how far they can be trusted and explains what they add to your field. Sheffield Hallam University’s research projects guide describes its purpose in much the same way. Put simply, the results chapter tells the reader what you found, while the discussion tells them why it matters.

The word itself carries two meanings. First, it means weighing results and their implications. Second, it means considering more than one side of an issue before reaching a position. Manchester’s Academic Phrasebank notes both senses, and strong chapters do both: they interpret each result, then test that interpretation against alternatives.

One rule holds almost everywhere. The chapter introduces no new findings, so every claim must trace back to data you already reported. If a point matters enough to discuss, make sure the evidence for it sits in your results.

Results, Discussion and Conclusion: Where Each Chapter Stops

Students often blur these three chapters, so a side-by-side view helps.

Aspect	Results	Discussion	Conclusion
Main job	Reports what the analysis found	Interprets and evaluates what the findings mean	Answers the main research question briefly and states the contribution
Voice	Descriptive and restrained	Analytical and argumentative	Brief and synthesising
Use of literature	Minimal	Heavy engagement with earlier studies	Little new engagement
New material	Your own data only	No new data, but new interpretation is expected	No new findings or evidence

A quick boundary test settles most doubts. If a sentence only describes data, it belongs in results. If it interprets data, it belongs in the discussion. If it answers your main research question as a whole, it belongs in the conclusion. For the first boundary, how to report your findings in the results chapter shows what to keep descriptive. At the other end, some programmes merge discussion and conclusion into one chapter, so confirm the format first, and use conclusion chapter support if you want a second opinion on how the two fit together.

Should Your Results and Discussion Be One Chapter?

Sometimes. Westminster’s dissertation structure guidance sets out three common layouts. A separate results chapter followed by a discussion chapter is typical of scientific and quantitative work. An analysis chapter followed by a discussion chapter, or a single combined analysis and discussion chapter, is more typical of qualitative work. If your course prescribes a structure, follow it.

Southampton’s guide to writing results and discussion adds that a combined chapter can still keep the two as distinct sections, with the discussion following on from the results. If you merge them, keep every interpretive move described below, but repeat the cycle theme by theme: state the finding, interpret it, then compare it with the literature before you move on.

Plan Your Discussion Section Before You Draft It

Start with a plain list of everything your results chapter reported, then decide which items you actually need to discuss. Findings that do not serve your aims cost word count and weaken focus, so cut them or give them a single sentence.

Next, build a three-column map. Each row pairs one finding with what earlier studies say and what the match or mismatch means.

Finding from my results	What the literature says	What the agreement or gap means
[Finding A, answers research question 1]	[Study or theory it supports]	[Your interpretation and its limits]
[Finding B, unexpected]	[Study it contradicts]	[Candidate explanations, ranked]
[Finding C, side result]	[Little or no earlier work]	[Decide: brief mention or cut]

Then rank the rows. Sheffield Hallam’s guide recommends leading with your most important or interesting findings, so rows that answer a research question directly should come first. If your questions feel too loose to answer cleanly, tighten them before you write: turning an idea into a research question shows how.

How to Write a Dissertation Discussion Chapter in Eight Parts

Not every dissertation needs eight separate sections, but the order below follows the logic markers expect: orient, answer, compare, explain, weigh, apply, qualify and point forward. Merge or rename parts to suit your department.

Part 1: Open with the aim and the headline findings

Remind the reader of your research problem, aims and questions, then summarise the key findings in one or two short paragraphs. Some examiners read the introduction and then jump straight to the discussion, so this reminder earns its place. Keep it tight: a statement of the overall result, not a replay of every table. Also reuse the key terms and tense from your opening chapter so both read as one project, and how to write a dissertation introduction shows which terms to fix early.

Part 2: Answer each research question in turn

State directly how your results answer each question. If you tested hypotheses, say whether the data support them. When a hypothesis fails, explain why that might be rather than treating it as a flaw, a point the University of Westminster library guide on findings, analysis and discussion makes as well. Hypotheses built from your literature review give every answer a clear target, which makes this part far easier to write.

Part 3: Compare your findings with the literature

Return to the studies in your literature review and sort your findings into three groups: those that agree with earlier work, those that contradict it, and those that add something earlier work did not address. Contradictions deserve the most space, because they are often the most interesting results. For each comparison, offer a reason, such as a different sample, method, context or period.

If a source you need only turned up after the review was written, ask your supervisor whether it should also appear there. A review that is purely descriptive makes comparison harder, so writing a literature review that sets up your comparisons is worth revisiting.

Part 4: Explain unexpected and non-significant results

Sheffield Hallam’s advice is to state the unexpected finding first, then offer your interpretation, then relate it to wider research. Non-significant findings still teach something, so do not bury them. Show that you considered measurement, sample, design and context before settling on an explanation, and phrase that explanation cautiously. Giving surprises their own sub-heading also stops them hiding inside longer paragraphs.

Part 5: Weigh alternative explanations

Research gathers evidence for ideas; it rarely proves them. Therefore ask what else could produce your pattern. Name one or two credible alternatives, say which your data favour and why, and admit where the data cannot decide. This habit separates a real discussion from a victory lap.

Part 6: State the implications for theory and practice

Ask what changes because of your findings. Do they support, challenge or refine a theory? Who could act on them, and how realistically? Keep claims proportionate to the evidence, since inflated claims and speculation beyond the data are among the quickest ways to lose marks.

Part 7: Acknowledge limitations without undermining the study

Cover only the limitations that bear on your objectives. Typical examples include a small or narrow sample that limits generalisability, problems during data collection, and variables you could not control. For each one, say how much it affects your conclusions, then restate why the findings still answer the question. Most limitations trace back to methodology choices you justified earlier, so refer to those choices rather than inventing new excuses. Keep the tone analytical and concise, never apologetic.

Part 8: Point to future research and close

Suggest specific next steps that follow from your limitations or surprises, instead of writing that more research is needed. A sentence or two is enough, because you are not drafting a second proposal. Some universities expect recommendations in the conclusion instead, so check where yours belong. Finish with a short synthesis that echoes your opening, so readers who skip ahead still meet the whole argument.

How to Organise a Discussion Chapter: Research Question, Theme or Theory

Structure should follow your argument, not your data files. Four organising principles cover most dissertations:

• By research question: the safest default, because it mirrors your aims and lets markers tick off each answer.
• By theme: suits qualitative work, and works best when it mirrors the themes in your results.
• By hypothesis: suits experimental and survey designs with clear predictions to confirm or reject.
• By theory or framework: suits studies where one framework drove the design and now needs evaluating.

Whichever you choose, put the strongest findings first and use consistent sub-headings so the argument is easy to follow.

Writing the Paragraphs: A Repeatable Pattern for Each Finding

Researchers who study discussion writing describe it as a cycle of moves. In a post drawing on John Swales’s work, the Thesis Whisperer lists four basics: restate the result, comment on it, evaluate it, and make a suggestion. Oxford Lifelong Learning’s discussion sections guidance adds that several claims build towards a key point, which the writer then states plainly. Combine the two and each finding becomes a short, repeatable paragraph:

1. Restate the finding in one sentence, pointing to the table, figure or theme in your results.
2. Comment: say whether it agrees with, contradicts or extends named studies.
3. Explain: give the most plausible reason and name one alternative.
4. Qualify: note any limitation that affects this particular finding.
5. Conclude the paragraph with one sentence stating what the finding means for your research question.

Calibrate your verbs as you go. Words such as “suggest” and “indicate” suit explanations and implications, while “prove” rarely survives scrutiny. However, Oxford’s guidance treats a key point stated in its own short main-clause sentence, without hedging, as stronger. Reconcile the two by hedging claims that stretch beyond your evidence and stating plainly those your data clearly support. The Academic Phrasebank groups sentence openers by move, so borrow the structure and rewrite the wording in your own voice.

How the Discussion Changes by Method and Discipline

• Quantitative work: interpret patterns, relationships and significant differences against your hypotheses, and say what the statistics mean for the real question rather than only reporting them.
• Qualitative work: interpret themes and support each claim with participant quotes or extracts, because assertion alone will not carry the argument.
• Mixed methods: explain how the strands fit together, for example by using qualitative findings as context for a quantitative pattern, before you interpret either.
• Across disciplines: Oxford’s examples show law and literature arguments moving through claims drawn from cases and texts, while a materials science example builds towards its key point from statements of result.

Common Discussion Chapter Mistakes and How to Fix Them

• Repeating the results. Revisit only the findings that carry your argument, and introduce each as a step towards a bigger point.
• Adding new findings. If a result matters, return to the results chapter and report it there first.
• Overclaiming. Replace “this proves” with wording that matches the strength of your evidence.
• Ignoring the literature. Tie every major finding to at least one earlier study or theory, even where you only partly agree.
• Hiding the surprise. Name the unexpected finding early, then interpret it.
• Apologising for limitations. State them accurately, size their impact, and restate the study’s value.
• Drifting from the research questions. If a paragraph does not help answer one, shorten it or remove it.

Final Checks Before You Submit Your Discussion

• Every claim traces back to data in your results chapter.
• Each research question and hypothesis has an explicit answer.
• Each major finding is compared with at least one source.
• Unexpected results get an interpretation and an alternative explanation.
• Limitations are specific, relevant and sized, not generic.
• Key terms and tense match your introduction.
• The opening and closing paragraphs would make sense to a reader who skips the middle.
• Your structure and tone resemble strong discussion sections in past dissertations from your own department.

Dissertation Discussion Chapter FAQs

Can I use “I” in my discussion chapter?

It depends on your department’s style rules. Oxford’s guidance treats a clear subject, such as “we” in natural sciences or “I” in social sciences and humanities, as a stronger way to state a key point than a vague “this suggests”. If first person is discouraged, use a specific noun such as “this study” instead.

Can I cite new sources in the discussion?

Most guidance expects you to compare your findings with the literature you reviewed. If an unexpected result sends you to a source you have not yet used, ask your supervisor whether it should be added to the review as well.

How long should the discussion chapter be?

No single figure applies everywhere. Word allocation depends on your institution, course and discipline, so check the module handbook and marking criteria. If the criteria reward critical analysis heavily, give interpretation the largest share of your words.

Next Steps for Your Draft

Put the writing on a calendar first. The dissertation timeline planner helps you reserve time for interpretation and redrafting, and the research question generator can sharpen any question that still feels vague.

If a draft stalls, discussion chapter support offers feedback on interpretation, literature comparison and structure. Model and reference work is intended for learning, so use it in line with your institution’s academic integrity policy.