How to start your dissertation: topic, proposal and timeline.

Step 1: Choose a topic that survives contact with reality

A good dissertation topic is:

• Narrow enough to cover properly in your word count
• Answerable with the sources and methods you can actually access
• Interesting enough to hold you for months
• Supervisable: someone in your department can guide it

Quick test: can you find ten relevant sources in an hour of searching? If not, narrow or change angle.

Step 2: Turn it into a research question

"Social media and mental health" is a topic. "How do first-year students at UK universities describe the effect of Instagram on study habits?" is a question. A good question names a population, an issue and a context, and can be answered with evidence rather than opinion alone.

Step 3: Write a one-page proposal

Most departments ask for something like this. Use your own handbook's format if it differs:

1. Working title
2. Research question (and any sub-questions)
3. Why it matters (the gap or debate)
4. Key sources or theories you will use
5. Method (what data or texts, how you'll analyse them)
6. Ethics considerations, if you involve participants
7. A rough timeline

Step 4: Build a timeline backwards

Take your submission date and work back. Example shape (adjust to your course):

• Final two weeks: proofread, check references, format, submit early
• Four to six weeks before: full draft finished, feedback from your supervisor
• Middle months: data collection or close reading, analysis, chapter drafts
• First weeks: topic, proposal, ethics approval if needed, reading list

Put the deadlines for ethics approval and any data collection at the front. Those are the ones that can't be moved. Our guide to organising your university work shows how to turn those dates into milestones you can track.

Step 5: Use your supervisor well

• Book regular meetings and send something short beforehand.
• Bring one specific question each time.
• Write a short note of what was agreed after each meeting.
• Ask what level of feedback your supervisor gives on drafts, and when.

Step 6: Where AI fits (and where it doesn't)

Check your department's rules first. If AI is permitted, it is most defensible for explaining unfamiliar concepts, checking that your plan is logical, or asking questions about your own draft. It should not write your analysis, invent sources or replace your reading. Always verify every reference against the real source.

Common early mistakes

• Choosing a topic too broad to finish
• Leaving ethics approval late
• Reading without taking structured notes
• Writing chapters in isolation without a clear argument. When you reach the writing stage, how to write a critical analysis and how to structure a university essay both apply at chapter level.

Frequently asked questions