How to Write a Dissertation

A dissertation is the longest, most independent piece of writing most students ever attempt, and the difficulty is rarely the writing itself. It’s the scale. Tens of thousands of words, months of work, and no weekly deadlines forcing progress. The students who struggle aren’t usually the worst writers; they’re the ones who treated a six-month project like a six-day one and tried to do it all in a final panic. This guide maps the structure and, just as importantly, the routine that gets you to the end.

Scope a question you can actually finish

Everything downstream depends on the research question, and the most common fatal mistake is choosing one too big for the time and word count you have. “How does social media affect mental health?” is a career, not a dissertation. “Is there a relationship between daily Instagram use and self-reported sleep quality among first-year undergraduates at one university?” is a project you can finish in a semester.

A good dissertation question is narrow, answerable with the access you actually have, and genuinely uncertain — there should be a real question, not a foregone conclusion. Before you commit, sanity-check it against three constraints: your word count, your timeline, and your access to data or sources. A brilliant question you can’t get data for is worse than a modest one you can.

Write a one-page proposal once you’ve settled it: the question, why it matters, and how you’ll investigate it. This page becomes your compass for the months ahead, and rereading it is the fastest way to catch yourself drifting off course.

Know the five-chapter map

Most empirical dissertations follow a five-chapter structure. You won’t write it in this order, but you need to know where everything goes:

1. Introduction. The question, why it matters, and what the dissertation will do. Written last, because you can’t introduce a study you haven’t finished.
2. Literature review. What’s already known, organized by theme rather than by source, ending in the specific gap your work addresses. This is an argument about the state of the field, not a list of summaries.
3. Methodology. Exactly what you did and why — your design, data, and analysis — in enough detail that someone could repeat it. The most mechanical chapter, which is why it’s the easiest to draft first.
4. Results. What you found, reported neutrally. No interpretation here; just the findings, with tables and figures doing the heavy lifting.
5. Discussion and conclusion. What the findings mean, how they answer your question, where they connect back to the literature, their limitations, and what should come next.

Humanities and theoretical dissertations often replace chapters three and four with thematic argument chapters, but the spine — question, context, your contribution, what it means — is the same. Check your department’s handbook for the exact expected structure; a stated requirement always overrides a general template.

Write in the order that’s easiest, not the order it reads

The finished dissertation opens with the introduction, but starting there is a trap — you’d be introducing conclusions you don’t have yet. Start with the methodology instead. It describes what you did, which is concrete and low-anxiety, and writing it early often clarifies gaps in your plan while you can still fix them.

A workable drafting order: methodology, then results, then literature review (now that you know exactly which prior work your findings speak to), then the discussion, and finally the introduction and abstract once the whole argument exists. This sequence means you’re always writing about something that already exists rather than staring down a blank page trying to predict the future.

Build a routine, because nothing else will force you

This is the part the structure diagrams leave out, and it’s where dissertations are won or lost. With no weekly deadlines, the work expands to fill — or rather, fails to fill — the months available. A routine replaces willpower with a schedule.

Two habits do most of the work:

• A fixed daily writing window. Even 90 minutes at the same time each day, treated as non-negotiable, compounds faster than occasional marathon sessions. Aim for a messy draft, not finished prose — you can edit a bad paragraph but not a blank one. Working out the right number of study and work hours a day keeps you productive without burning out months before the deadline.
• A backward-planned schedule. Start from the submission date and work backwards, assigning chapters to weeks and leaving real buffer for the things that always go wrong — data that takes longer to collect, an analysis that needs redoing, a supervisor who’s away for two weeks. The study schedule guide covers how to build a plan that survives contact with reality. A dissertation timeline with no slack is a timeline that will break.

Break each chapter into section-sized tasks on your schedule. “Write literature review” is paralyzing. “Draft the 600 words on prior work about sleep and screen time” is a task you can finish today and tick off, and the ticking matters more than it sounds.

Manage your sources before they manage you

A literature review can involve dozens or hundreds of papers, and disorganized reading is how months disappear. From the very first source, record the full citation, the core claim, and one line on how it relates to your question. A reference manager is worth setting up on day one.

To get through the reading volume, read strategically rather than cover to cover. A notes summarizer can compress a long paper to its central claims so you can triage which deserve a deep read, and turning the key definitions and concepts you must retain into flashcards lets spaced repetition keep them in memory across a months-long project — so you’re not re-reading the same foundational papers every time you sit down. The thinking and synthesis stay yours; the tools just stop the reading from swallowing the calendar.

Revise in layers and leave time for it

A dissertation can’t be fixed in one read-through, so revise in deliberate layers, ideally with a day or two between them:

• Argument layer. Does each chapter do its job, and do they connect into one coherent story that answers your question? This is the level supervisors care about most.
• Chapter layer. Within each chapter, is the order logical and does every section earn its place?
• Sentence layer. Clarity, signposting, and flow — read sections aloud to catch what your eye skips.
• Mechanical layer. Citations, formatting, figure numbering, and the reference list. Tedious, and a reliable source of lost marks if rushed.

Budget a couple of weeks for revision and formatting alone. Students routinely underestimate this and submit a strong analysis wrapped in a sloppy, inconsistent document — the academic equivalent of a good meal served on a dirty plate.

The honest summary: a dissertation is not hard because the writing is hard. It’s hard because it’s long, self-directed, and easy to put off. Beat it by scoping a question you can finish, writing in the order that’s easiest, and protecting a daily routine that makes steady progress the default instead of the exception. Do that, and the longest thing you’ll write becomes a series of ordinary days rather than one impossible deadline.

FAQ

How do I start writing my dissertation?

Start by narrowing to a research question you can answer with the time and access you actually have, then write a one-page proposal: the question, why it matters, and how you'll investigate it. From there, the easiest chapter to draft first is usually the methodology, because it describes what you did rather than what you concluded. Don't start with the introduction — you can't introduce a study you haven't done yet, so most people write it last.

What are the 5 chapters of a dissertation?

The classic empirical dissertation has five chapters: (1) Introduction — your question and why it matters; (2) Literature Review — what's already known and the gap you're filling; (3) Methodology — how you gathered and analyzed your data; (4) Results — what you found, reported neutrally; and (5) Discussion and Conclusion — what the findings mean, their limits, and what comes next. Humanities dissertations often use thematic chapters instead, but the underlying logic is the same.

Can you write a dissertation in 2 weeks?

Writing up an analysis you've already completed in two weeks is hard but possible if you have your data and reading done. Doing the actual research — collecting data, reading the literature, and writing — in two weeks is not realistic for a quality result. If you're under extreme time pressure, narrow your scope aggressively, write the methodology and results first since they're the most mechanical, and accept that a finished, defensible dissertation beats an ambitious unfinished one.

How long is a dissertation?

It varies widely by level and field. An undergraduate dissertation is often 8,000 to 12,000 words, a master's dissertation 15,000 to 25,000, and a PhD thesis can run 70,000 to 100,000. Your department sets the real number, so check your handbook — the word count shapes how you scope your question, and a question too big for your word limit is the most common planning mistake.

What is the difference between a thesis and a dissertation?

The terms are used differently across countries. In the UK, a dissertation is usually the long project at undergraduate or master's level and a thesis is the PhD document. In the US, it's often reversed — a thesis at master's level and a dissertation for the PhD. The structure and standards are similar; what matters is the convention your own institution uses, so follow your department's terminology.

How do I stay motivated writing a dissertation?

Treat it as a series of small, scheduled tasks rather than one enormous document. Break it into chapter-sized and then section-sized pieces, give yourself a fixed daily writing window, and aim for a messy draft rather than perfect prose — you can't edit a blank page. A realistic schedule with built-in buffer days protects momentum better than motivation does, because it removes the daily decision of whether to work.