How to choose a dissertation topic you can actually finish
The best topic is not the most interesting one in the abstract. It is the most interesting one you can answer properly with the time, data and skills you really have.
Start from a question you care about
You will live with this topic for months. Begin with the modules, debates or real-world problems that held your attention, and write down the questions they left you with. A topic you are curious about survives the dull weeks of data collection; one chosen because it sounded impressive usually does not.
At this stage a broad area is fine — "mobile money", "exam stress", "urban flooding". The work of the next steps is to turn an area into a question.
Test it against the four things that sink projects
Most projects that fail do not fail on ideas. They fail because the data could not be obtained, the scope was too big for the word count, ethics approval took too long, or the method needed skills the student did not have. Check each of those before you commit.
Data access: who or what would you study, and can you realistically reach them? A survey of 200 hospital patients needs permissions that a survey of 40 classmates does not. Scope: could you answer it well in your word count, or only superficially? Ethics: does it involve minors, patients, or sensitive topics that need a longer approval process? Skills: does it need statistics, coding or a language you would have to learn first?
Narrow the area into a question
Narrow along the dimensions a reader would ask about: who (which group), where (which place), when (which period), and what exactly (which effect or relationship). "Mobile money" becomes "How has mobile money changed how women market traders in one town manage stock?" — something a single student can investigate and answer.
A good test: could someone else read your question and know what evidence would answer it? If not, it is still a topic, not a question.
Check there is something to find
Search your library database and a scholarly search engine for your narrowed question. Too little existing work can mean there is nothing to build on; a settled answer means there is nothing left to find. What you want is a conversation you can join — studies that leave a gap, disagree, or were done somewhere else.
Checklist
- I can state my topic as one question, not a subject area.
- I know who or what I would study, and I can reach them.
- It fits my word count and deadline with time left for writing.
- I know whether it needs ethics approval, and how long that takes.
- I have, or can learn in time, the methods it needs.
- I have found existing work to build on — and a gap in it.
- I still find it interesting after thinking about the dull parts.
Common mistakes
- Choosing the topic before checking whether the data exists.
- Keeping the scope national or global when one town, school or company would give a better answer.
- Asking a yes/no question, which leaves nothing to analyse.
- Picking a topic to impress rather than one you can finish.
- Leaving ethics approval until after the method is designed.
Get a supervisor's eye on your own project
Tell the StudyAI Supervisor your interests and constraints and it checks each idea for data access, scope, ethics and more — or suggests directions if you have none yet.
Frequently asked questions
How do I know if my dissertation topic is viable?
Check it against data access, scope, ethics, skills and time. If you can reach the people or data, answer it properly within the word count, get approval in time, and use methods you have or can learn, it is viable.
How narrow should a dissertation topic be?
Narrow enough that one person can answer it well in the word count: usually one group, one place and one period. Broad topics produce description instead of analysis.
Can I change my topic after starting?
Usually, early on — most supervisors expect the question to sharpen. Check your department's rules on deadlines for changing the approved title.