What each chapter has to establish, why the discussion is where most theses are weakest, and how to prepare for the viva.
A thesis is an argument, not a report of everything you did. Each chapter has one job: to establish something the next chapter depends on. Most theses are strongest in the methods and weakest in the discussion, because the discussion is the only chapter that cannot be written by describing what happened — it requires you to say what your findings mean and why that matters, which is exactly what the examination is about.
Worth clearing up, because the terms are reversed between countries and the confusion is genuine.
In the UK and most of the Commonwealth, a thesis is the doctoral document and a dissertation is usually the shorter master’s one. In the United States it is generally the other way round: a dissertation is the doctoral work, a thesis the master’s. Elsewhere usage varies, and some institutions use them interchangeably.
Nothing about the work changes. Use whichever term your own university uses, and take the requirements from your institution’s own regulations rather than from general guidance, since word limits, structure and examination arrangements vary considerably.
Structures vary by discipline — some fields expect papers stapled into a frame rather than a monograph — but in the conventional shape each chapter carries a specific burden. A chapter that only describes has not carried it.
| Chapter | Its job | How it usually fails |
|---|---|---|
| Introduction | Establish that the problem matters and state what the thesis will show | Background with no argument; the contribution never stated |
| Literature review | Show where the field’s account runs out, and that your question follows | Studies summarised in sequence; no position taken |
| Methodology | Justify the design as the right way to answer that question | Textbook recital; methods listed rather than reasoned |
| Results / Findings | Report what was found, clearly and without interpretation | Interpretation smuggled in; output presented table by table |
| Discussion | Say what the findings mean, against the literature | Findings repeated in different words |
| Conclusion | State the contribution and its limits | A summary of the thesis instead of a claim |
The test to apply to any chapter: what does the reader know at the end that they did not know at the start, and does the next chapter need it? If you cannot answer, the chapter is description.
This is where most theses are weakest, and the fix is largely structural.
Findings report what you found: what the data show, presented so a reader could reach their own view. No comparison with other studies, no explanation of causes, no significance claimed beyond what was measured.
Discussion does the interpretive work: what the findings mean, how they sit against existing literature, where they agree and where they conflict, why that might be, what follows for theory and practice, and what the study cannot say.
Two failure modes recur. The first is a discussion that restates findings in different words — recognisable because it cites almost nothing, when a discussion should be dense with references as your results are placed against the field. The second is findings contaminated with interpretation, which leaves the discussion nothing to do and makes it hard for a reader to separate what you observed from what you concluded.
A discussion that is working usually contains a paragraph on something that did not come out as expected, and takes it seriously rather than explaining it away. That paragraph is often what convinces an examiner you understand your own study.
A doctorate requires an original contribution to knowledge, and candidates are routinely unable to state theirs in a sentence — which is a problem, because it is a standard opening question in a viva.
Originality does not have to mean a new theory. It can be a question nobody has asked, evidence from a context where it was missing, a method applied where it has not been, a synthesis that reorganises how a field is understood, or a finding that complicates an accepted account. Modest and precise beats grand and vague.
Try writing it as a single sentence: this thesis shows that…, which matters because… If the sentence needs three clauses of context before it starts, the contribution is not yet clear to you either. That sentence, once you have it, belongs in the introduction, the discussion and the conclusion — and it is worth being able to say it out loud without notes.
Limitations sections are commonly written as a list of regrets, which reads as weakness. A strong limitations section reads as command of your own design.
The difference is in what you do with each one. “The sample was small” is an apology. “The sample supports claims about how these practitioners understood the policy, but not about how common that understanding is, which would require a survey design” is command — it names the boundary, explains what it rules out, and shows you know what would be needed to go further.
Cover the ones an examiner will raise anyway: sampling and who was excluded, design constraints on the claims, measurement limits, your own position where it shaped the work, and anything about the context that limits transfer. Raising them yourself is considerably better than being led to them.
The viva is not a hostile examination of whether you deserve the degree. It is a conversation to establish that the work is yours, that you understand it, and that you can defend the choices in it. Candidates who struggle are usually those who have never articulated their argument aloud.
What is worth preparing:
A mock viva with someone who has examined doctorates is the single most useful preparation available, because it surfaces the questions you have not thought about while there is still time to think about them. Saying “that is a fair point, and I would address it by…” is a perfectly good answer; pretending a weakness is not one is not.
It depends where you are. In the UK and much of the Commonwealth, a thesis is the doctoral document and a dissertation the master’s one; in the United States it is typically reversed. Some institutions use the words interchangeably. Follow your own university’s regulations, which is also where the binding requirements on length and structure come from.
Findings report what the data show, without interpretation. Discussion says what that means: how it sits against the literature, where it agrees or conflicts and why, what follows, and what the study cannot claim. A discussion that restates findings in different words is the most common structural weakness in doctoral writing, and it is recognisable because it cites almost nothing.
Originality can be a new question, evidence from a context that lacked it, a method applied somewhere new, a synthesis that reframes a field, or a finding that complicates an accepted account — it does not have to be a new theory. The practical test is whether you can state it in one sentence of the form “this thesis shows that…, which matters because…”. If that sentence is hard to write, the issue is usually clarity rather than insufficiency.
Name each limitation, say precisely what it rules out, and say what would be required to go further. That turns the section from a list of regrets into evidence that you understand your own design. Cover what an examiner would raise anyway — sampling, design constraints on causal or general claims, measurement, and your own influence on the work — because raising them yourself is far stronger than being led to them.
Be able to state your contribution in a sentence without notes, know why you chose your design over the obvious alternative, and have an honest account of your weakest point ready. Read what your examiners have published on your topic, particularly where your work disagrees with theirs. Then have a mock viva with someone who has examined doctorates — saying the argument out loud is what most candidates have never done before the day itself.
Proofreading and language editing are ordinary parts of academic work, and most universities publish guidance on what is acceptable — typically correcting language, clarity and consistency, while leaving the substance and the argument as yours. Check your own institution’s policy, since some ask that editorial help be acknowledged, and follow it. Doing so is straightforward and removes any question later.
A draft read by someone who has examined doctorates surfaces the questions you will be asked while there is still time to answer them. Describe where you are and what is not landing.
Leave an email or a WhatsApp number — whichever you prefer — and tick how we should reach you. We reach out within 30 minutes.
