A practical, step-by-step guide to writing a dissertation, covering topic selection, proposals, structure, research, writing, and editing.
What a dissertation actually is
A dissertation is a long, independent research project that asks a specific question and answers it using evidence โ not a longer version of a regular essay. It's the piece of work most degree programs use to test whether you can design a study, carry it out, and argue for your conclusions on your own.
That independence is what makes it feel different from other assignments. There's no weekly reading list telling you what to think about next; you're setting the direction yourself, usually with a supervisor checking in periodically rather than teaching you the material.
Step 1: Choose a topic you can actually finish
The best dissertation topic isn't the most ambitious one โ it's the one you can complete with the time, data, and access you actually have. A brilliant idea that needs six months of fieldwork you can't get funding for is a worse topic than a modest one you can execute well.
Start broad, then narrow. Pick a general area you're genuinely interested in, read five or six recent papers or books in it, and look for a specific gap: a question nobody has quite answered, a method nobody has applied to this context, or a debate where the evidence is thin. Run your narrowed idea past your supervisor before committing โ they'll often spot feasibility problems you can't see yet.
Signs a topic is workable:
- You can state your research question in one sentence
- The data or sources you need are realistically accessible
- Existing literature exists to build on, but hasn't fully answered your question
- You can explain in two sentences why the answer matters
Step 2: Write the proposal
Most programs require a dissertation proposal before you start, and even if yours doesn't, writing one anyway is worth the time. A good proposal forces you to think through problems on paper instead of discovering them three months into data collection.
A typical proposal includes your research question, a short review of relevant existing work, your planned method, and a rough timeline. Keep it honest about what you don't know yet โ supervisors expect a proposal to have open questions, not a finished argument.
| Proposal section | What it should answer |
|---|---|
| Research question | What exactly are you trying to find out? |
| Background | Why does this question matter, and what's already known? |
| Method | How will you answer the question โ survey, interviews, experiment, archival analysis? |
| Timeline | What happens in which month, roughly? |
| Limitations | What can't this study do, and why is that acceptable? |
Step 3: Understand the typical structure
Dissertation structures vary by field, but most empirical dissertations in the sciences and social sciences follow a similar skeleton. Humanities dissertations are often organized more thematically by chapter, but the introduction, literature review, and conclusion still play the same roles.
| Chapter | Purpose | Typical share of word count |
|---|---|---|
| Introduction | States the research question and why it matters | 5โ10% |
| Literature review | Shows what's already known and where the gap is | 15โ25% |
| Methodology | Explains how you answered the question | 10โ15% |
| Results/Findings | Presents what you found, without interpreting yet | 20โ25% |
| Discussion | Interprets the findings and connects them back to the literature | 15โ20% |
| Conclusion | Summarizes contribution, limitations, and future work | 5โ10% |
Check your own department's handbook before relying on this table โ some fields combine results and discussion into one chapter, and word-count splits differ.
Step 4: Build the literature review as you go
Treat the literature review as an ongoing document from day one rather than a chapter you write once at the end. Every time you read something relevant, add a few sentences on what it argued, how it relates to your question, and where you'd place it in your emerging argument.
A literature review isn't a summary of everything you read โ it's an argument about the state of the field that justifies your own research question. Group sources by theme or debate rather than listing them chronologically or one paper at a time, and be explicit about where the gap in existing knowledge sits.
Common literature review mistakes:
- Listing sources one after another with no synthesis between them
- Summarizing instead of critically engaging with weaknesses or disagreements
- Only citing work that supports your position, ignoring contradicting studies
- Leaving the gap your dissertation fills unstated or vague
Step 5: Collect and analyze your data
Your methodology chapter should be detailed enough that someone else could repeat your study from reading it alone. Record decisions as you make them โ sample size, tools used, exclusion criteria โ rather than trying to reconstruct your reasoning months later while writing up.
When analyzing data, separate the results chapter from the discussion chapter, even if your final structure eventually merges them. Report what you found first, without interpretation, then interpret it afterward. This keeps your findings honest and stops the temptation to shade results toward the conclusion you expected going in.
Step 6: Write in stages, not start to finish
Very few dissertations get written in chapter order. Most experienced writers draft the methodology and literature review early, since those depend less on final results, and write the introduction and abstract last, once they know exactly what the dissertation actually argues.
Set a rough weekly word or page target rather than waiting for large blocks of uninterrupted time. A dissertation is long enough that steady, moderate progress beats sporadic bursts โ 500 written words most days adds up to a full first draft faster than waiting for a free weekend that rarely comes.
A simple staged approach:
- Draft methodology and literature review first
- Collect and analyze data
- Write results/findings
- Write discussion, connecting findings back to the literature review
- Write a working introduction and conclusion
- Revise the introduction and abstract last, once the full argument is settled
Step 7: Edit in separate passes
Editing a dissertation in one pass usually misses things, because you're trying to check argument, structure, and grammar all at once. Split editing into distinct passes instead.
Do a structural pass first โ does each chapter flow logically, does the discussion actually answer the research question, are there gaps in the argument. Then a clarity pass, checking that each paragraph makes one clear point. Save line-level grammar and formatting checks for last, once the content itself is settled, so you're not proofreading sentences you'll end up cutting.
Common mistakes to avoid
Scope creep is the most common problem โ a topic that keeps expanding as you research it, until the original question gets buried under three new ones. Revisit your original research question regularly and cut anything that doesn't directly serve it.
Weak transitions between chapters make a dissertation read like several disconnected papers stapled together. Each chapter should end by pointing toward the next, and each new chapter should briefly reconnect to what came before.
Underestimating the editing timeline is another frequent issue โ writers often budget time to finish a first draft but not to revise it properly, then rush the final weeks. Build at least three to four weeks of editing time into your schedule, separate from writing time.
Conclusion
A dissertation is manageable when it's broken into stages: a workable topic, a clear proposal, a structure you understand before you start writing, and an editing process that happens in deliberate passes rather than all at once. Treat it as a long project built from small, steady progress rather than a single act of writing, and the scale of it becomes far less intimidating.
Frequently asked questions
How long does it take to write a dissertation?
Most dissertations take between six months and a year to complete, depending on the degree level, research method, and how much data collection is involved. Undergraduate dissertations are usually shorter and faster; PhD dissertations can take several years alongside ongoing research.
How long should a dissertation be?
Length varies by institution and level. Undergraduate dissertations are often 8,000 to 15,000 words, master's dissertations 15,000 to 25,000 words, and PhD dissertations commonly run from 60,000 to 100,000 words. Always check your department's specific requirements first.
What's the difference between a dissertation and a thesis?
In many countries the terms are used interchangeably, but in the US, a thesis usually refers to a master's-level project and a dissertation to a doctoral one. In the UK, it's often the reverse, so check the terminology your institution uses.
Can I change my dissertation topic partway through?
Yes, though it's easier the earlier you do it. Minor adjustments to scope are normal as research develops. A full topic change after data collection has started usually requires approval from your supervisor and may affect your timeline.
Should I write the introduction first or last?
Most experienced writers draft a working introduction early to guide their research, then rewrite it properly at the end once the full argument and findings are clear. The final introduction should accurately reflect the dissertation you actually wrote.
How many sources should a dissertation cite?
There's no fixed number โ it depends on your field and topic. What matters more than quantity is that your literature review covers the key work in your area and that every claim you make is properly supported.