Most supervision meetings about topics are over in ten minutes. The ones that end with a yes have one thing in common: the student walks in with a question, not a subject area.
Finding a topic for your dissertation is a process, not a flash of inspiration. You collect ideas widely. Then you narrow each idea down by time frame, place, target group and angle until exactly one research question is left. Finally you check whether that question can be answered with the data and the weeks you actually have. What you end up with is a topic your supervisor will accept and that you can finish before the deadline.
📌 The key points at a glance
- A good topic leads to exactly one research question.
- Narrow it down by time frame, place, target group and angle.
- One UK credit stands for roughly 10 notional learning hours.
- A 20-credit module therefore adds up to about 200 hours.
- No empirical topic survives without realistic access to data.
Create a survey for free
With empirio.ai you can create a modern online survey in minutes — with hosting in the EU.
- AI-built survey
- Adjust by drag & drop
- Real-time analysis
What makes a good dissertation topic
A good dissertation topic meets four conditions at once: it holds your interest for months, it fits your degree subject, it can be turned into exactly one research question and it can be finished in the time you are given.
Most students only test the first condition. Interest matters most, because a topic that bores you after four weeks will not carry the remaining two months. Interest on its own is not enough though. Nobody marks your work on how exciting the field is. Markers look at whether one question was answered properly.
The combination is what counts. A topic that thrills you but yields no answerable question ends up as an essay. A topic with a perfect question and no way of getting data ends up as a rescue job three weeks before submission.
| Condition | How you recognise it | What happens without it |
|---|---|---|
| Interest | you keep reading without forcing yourself | the work stalls in week four |
| Subject fit | a module on your programme covers it | no supervisor feels responsible |
| Answerability | you can write one single question | the text turns into a pile of material |
| Feasibility | data and time are both available | the method changes just before the deadline |
Whether your dissertation needs data of your own or can work from existing literature is worth deciding right here. The differences and what they mean for your workload are set out in the guide on choosing between an empirical and a theoretical dissertation.
Step 1: collect ideas until you can choose
Ideas for your topic come from five reliable places: the modules you enjoyed, the limitations sections of recent journal articles, your placement or part-time job, the research groups in your department and ready-made topic lists.
Limitations sections are the most underrated of the five. Almost every study ends with a paragraph in which the authors write down what their work could not do and what someone should look at next. Researchers have handed you a research gap in finished sentences. Ten minutes of reading is all it costs you.
Collect widely at this stage and judge nothing yet. Ten to fifteen ideas make a solid starting pool and you will cross most of them out again.
- Modules and coursework: topics you already know something about.
- Limitations and further research: open questions at the end of recent studies.
- Placement or part-time job: problems you noticed at work.
- Research groups in your department: live projects and specialisms.
- Topic lists: prompts to think further, not to copy.
If you would rather start from a ready-made selection, there are topic suggestions for bachelor’s and master’s dissertations sorted by subject area. This guide describes the route from idea to research question, that list supplies the raw material. Should you want to write with a company, the particulars of a dissertation written at a company are worth a look, because the topic often arrives ready-made.
Keep your ideas and questions in a table
A table beats a list for this collection, because three columns show you at a glance which idea already yields a question and which does not. Fill the third column later, in step 3.
| Topic | Possible research question | Feasibility notes |
|---|---|---|
| Second-hand clothing apps | How do young shoppers judge resale apps? | survey, reach through the students’ union |
| Hybrid working and loyalty | How does hybrid working relate to staying with an employer? | HR has to agree first |
If an hour of searching produces no question for an idea, cross it out. The middle column exists for exactly that decision.
Our reading tip
Bell, Judith and Waters, Stephen (2018): Doing Your Research Project. A Guide for First-time Researchers. 7th edition, London: Open University Press. The book walks first-time researchers from the opening idea through to the finished project and is the model for the table above.
Step 2: narrow the topic until one research question is left
You narrow a topic by shrinking it along four settings one after another: time frame, place or institution, target group and angle. “Sustainable fashion” turns into a question you can ask in one sentence and answer with data of your own.
The mistake almost everyone makes is narrowing in one jump. “Sustainable fashion” becomes “resale apps in Manchester” straight away and the target group, the time frame and the perspective are all still missing. Turn one setting at a time and check after each step whether the topic still holds up.
A simple rule closes this step: you are done when you can no longer say your topic in three words without leaving out something important.
| Setting | Question to your topic | Example |
|---|---|---|
| Time frame | Which period? | since resale apps reached the UK high street |
| Place or institution | Where exactly? | Manchester |
| Target group | Who is being looked at? | students aged 18 to 24 |
| Angle | From which perspective? | what they think, not what they buy |
All four settings together produce a question like this one: “How do students aged 18 to 24 in Manchester judge second-hand clothing apps?” Not a beautiful sentence, but a workable one. Workable is the only thing that matters at this point.
As long as you can say your topic in three words, it is not yet a topic.
Turning that narrowed topic into a clean question is set out step by step in the guide on how to write a research question. The final title of your dissertation is something else again and follows its own rules: wording the title of a bachelor’s or master’s dissertation.
Step 3: check feasibility before you register your topic
A topic is feasible when you can answer its research question with data you can genuinely reach inside the submission window. Credits give you a first estimate of the hours involved.
The QAA credit framework describes one credit as equal to 10 hours of notional learning, so a 20-credit module stands for around 200 hours of work. A full-time year carries 120 credits and a bachelor’s degree with honours 360. You can read the wording in the Higher Education Credit Framework for England (second edition, 26 May 2021).
One thing the framework does not do is fix the size of a dissertation. No national rule says how many credits it carries. Universities decide that themselves and the framework is advisory rather than regulatory in any case. Your course handbook and your programme regulations are what count, and only they tell you how many weeks you really have.
A quick sum
A 40-credit dissertation works out at roughly 400 notional hours. Set aside 80 of those hours for recruiting participants and analysing results. That leaves around 320 hours for reading, writing and proofreading. Do this sum before you register your topic, not afterwards.
Empirical feasibility: can you reach enough data?
The usual breaking point of an empirical dissertation is access to the field, not the statistics. A survey inside a company needs sign-off from the company. Many UK universities also expect ethics approval from a research ethics committee before you collect anything from people, and your department’s research ethics policy sets out what applies to your project. Both steps take weeks.
How many responses you need depends on what you want to do with them: describing one group takes fewer than comparing two. Methods and sizes are covered in the guide on the sample. Where to find respondents without asking the same friends three times is covered in finding survey participants.
Checklist before you register your dissertation topic
Work through these six points before your topic becomes binding. Any point you cannot answer with yes is an open item rather than a detail that will sort itself out later. Point four is where it usually sticks.
- The research question fits into one sentence.
- Your supervisor has said yes to the topic out loud.
- Every term in the question has literature behind it.
- Access to the field is settled, ethics approval included.
- The method follows from the question, not the other way round.
- Fieldwork and analysis fit into your timetable.
Create a survey for free
With empirio.ai you can create a modern online survey in minutes — with hosting in the EU.
- AI-built survey
- Adjust by drag & drop
- Real-time analysis
Finding a dissertation topic with AI: what helps and what does not
Language models are good at turning one broad field into twenty variations and poor at reporting the state of research correctly. Use them to sort and sharpen ideas you already have, never as evidence that enough literature exists on a topic.
In practice it works like this. You give the model your field and your programme and ask it to narrow the field along the four settings from step 2. Nothing about that replaces a literature search, but it does break the blank-page deadlock. Every idea that comes out gets the same scrutiny as one of your own.
Whether you may use AI at all and whether you have to declare it is for your university to decide. Many programme regulations and academic integrity policies now say so explicitly, others still say nothing. Ask your supervisor before you start rather than when you sign the declaration of authorship.
Careful
Language models invent references that look plausible and do not exist. Check every title in your library catalogue or a subject database before it goes into your proposal. If you cannot find it there, it does not exist for your dissertation.
Ready-made tools for this step exist as well, for instance the AI section of GoThesis, which belongs to the same company as empirio.ai. The rule above still applies there: suggestions are a starting point and the checking stays with you.
Common mistakes when choosing a topic
Most topics fail not because they are bad but because a step was skipped. Four patterns come up again and again.
Read the following four sections before you register your topic. Afterwards every correction costs time you will want for writing.
The topic stays too big for the time available
“Artificial intelligence in HR” sounds like a topic and is a field with several thousand publications behind it. Register that and you spend the first weeks on literature that never makes it into the text, then narrow down under time pressure. What comes out is usually a dissertation that touches many things and answers none.
The supervisor is approached too late
Many students look for the perfect topic first and for someone to supervise it second. Supervision places are limited and tied to specialisms, so that order works against you. Talk to two or three academics early about a rough topic and let them tell you what fits their area. A no in week one costs you nothing. A no in week four costs you your topic.
Data collection gets planned after registration
Planning fieldwork late is the most expensive mistake and the least visible one, because everything looks fine right up to registration. Only afterwards does it emerge that the company will not allow a survey, that ethics approval takes six weeks or that ten days have produced fourteen responses. Sort out access, approvals and your distribution plan while you are still choosing the topic.
The whole topic gets swapped at the first problem
When fieldwork stalls, the first instinct is usually the biggest lever: a new topic. Yet the fault almost never sits in the topic. One setting is normally all that needs adjusting. Shrink the target group, move from a survey to five interviews, shift the time frame or limit yourself to one institution. Saving a topic takes days. A new topic costs weeks.
Conclusion
Choosing a topic is not a matter of inspiration but a sequence of three decisions: collect, narrow, check. Anyone who brings the third decision forward and settles where the data will come from before registering is spared the most unpleasant experience of the whole degree. And once you can put your topic into a single sentence that ends in a question mark, you are further along than most people in your year!
What comes next
- Want to see the whole sequence? Empirical research process: phases, steps and example
- Writing your first dissertation? Bachelor thesis: guide and structure with an example
- Need the order of the chapters? Dissertation structure: the order of sections explained
Topic settled, now you need data?
With empirio.ai, an online survey tool from Germany, you create the questionnaire for your dissertation free of charge, share it as a link or QR code and see your responses analysed straight away.