The topic you pick in week one decides how hard the next six months will be. Choose something you cannot get data for, and no amount of writing skill rescues it. Choose something too broad, and you will still be narrowing it in March. Most students do not need more ideas — they need a way to tell which of their ideas is researchable.
So this page does two things: it gives you starting points across nine UK subject areas, and it gives you the four tests every one of them has to pass before you commit. If you already have your question and need to get the writing moving, you can start drafting in Tesify.
The four tests: run these before you fall in love with an idea
- Can you get the data in the time you have? Not “does the data exist” — can you obtain it, this term, with your access and approvals. This kills more topics than anything else.
- Is there enough literature, but not too much? Too little and you cannot build a case; too much and you cannot say anything new. If a search returns a handful of relevant papers, that is roughly the right density for an undergraduate project.
- Can you state it as one question? If your topic needs three sentences to explain, it is a field, not a question. “What do students think about AI?” is a field. “How do final-year students at one UK university describe deciding whether to use AI for coursework?” is a question.
- Will your department approve it? Some topics need ethical clearance you cannot obtain — health research involving NHS patients being the clearest example. Check feasibility before design, not after.
Psychology
- Does the format of revision (self-testing versus rereading) predict exam confidence independently of actual performance?
- How does perceived control over working hours relate to burnout among student part-time workers?
- Are self-reported sleep quality and academic self-efficacy associated in final-year undergraduates?
- Does exposure to curated social media content relate to body image concerns differently by gender?
- How do students describe the experience of imposter feelings during dissertation writing? (Qualitative)
Psychology projects live or die on the measurement decisions. Use validated scales wherever they exist — they bring published reliability evidence with them and save you defending an instrument you wrote yourself.
Business and management
- How do UK SMEs in one sector describe the trade-offs of adopting four-day working?
- Does employer branding on recruitment pages differ between sectors competing for the same graduates? (Content analysis — no participants needed)
- How do small hospitality firms manage seasonal staffing, and what do managers see as the main constraint?
- What sustainability claims appear in annual reports of listed UK retailers, and how have they changed over five years?
The content-analysis options here are worth a hard look. Annual reports and public web pages are free, plentiful, need no recruitment, and usually attract minimal ethical review.
Law
- How have UK courts approached a specific statutory test since a named leading case?
- Does the current regulatory framework in one area adequately address a specific gap, and what would reform require?
- A comparative analysis of how two jurisdictions treat the same legal problem.
- How do published sentencing remarks in one offence category reflect stated guideline factors?
Law dissertations are doctrinal far more often than empirical, which removes the data-collection problem entirely and replaces it with a different discipline: your argument has to be built from primary sources — legislation and case law — rather than from commentary about them.
Education
- How do trainee teachers describe managing workload during placement? (Qualitative, with careful attention to recruiting through gatekeepers)
- Is there an association between school-level characteristics and attainment measures published by the Department for Education? (Secondary data)
- How is a specific curriculum policy interpreted differently across published school policies?
- What do published inspection reports say about a particular practice, and how consistently?
Sociology and social policy
- How does housing tenure relate to reported wellbeing in a national survey dataset?
- How do young adults describe the transition out of full-time education in a period of high housing costs? (Qualitative)
- Does participation in community organisations vary by area deprivation?
- How is a specific social issue framed across UK newspapers of different political leanings? (Media content analysis)
Nursing, midwifery and allied health
Read this section before choosing, because the constraint here is unusually hard. The Health Research Authority states that standalone research at undergraduate level requiring ethics review or HRA approval cannot take place, and that undergraduate applications are no longer accepted for Research Ethics Committee review. A supervisor as chief investigator does not change this for a standalone project.
Viable routes instead:
- A literature-based review of the evidence for a specific intervention.
- Analysis of published NHS England statistics on service use, waiting times or outcomes.
- Analysis of publicly available policy or guidance documents.
- Research on a health topic that does not involve patients, service users or NHS staff as participants — for example, public health knowledge among a non-clinical population.
These are the HRA’s own suggested alternatives, not workarounds, and they produce perfectly strong dissertations. The full restriction is set out in our guide to survey platforms and ethical approval.
Criminology
- How does recorded crime of a specific type vary across neighbourhoods, and what area characteristics correlate with it?
- Do stop and search outcome rates differ across police force areas? (Open data, no approvals needed)
- How is a particular offence category represented in news coverage compared with recorded incidence?
- What do published force policies say about a specific practice, and how much do they diverge?
Geography and environmental science
- How has land use changed in one area, and what does the pattern suggest about policy?
- Does access to green space vary with deprivation across a city’s wards?
- How do residents describe experiencing flood risk in an affected community? (Qualitative)
- What spatial patterns appear in a published environmental monitoring dataset?
Computer science
- Build and evaluate a tool that solves a specific, narrow problem, with a defined evaluation method.
- Compare the performance of approaches to one task on a public benchmark dataset.
- A usability evaluation of an existing system against recognised heuristics.
- An analysis of a security or accessibility property across a defined set of public websites.
Computer science projects need their evaluation designed at the start. “I built it and it works” is not a finding; “I built it and here is how I measured whether it works, against what baseline” is.
How to narrow a topic in four moves
Take any idea above and apply these in order.
- Add a population. “Student wellbeing” becomes “wellbeing among final-year undergraduates”.
- Add a context. “…at one UK university” or “…in the hospitality sector”.
- Add a relationship. Name the two things you think connect: “…and self-reported sleep quality”.
- Add a method. Decide whether you are measuring an association or exploring an experience, because that determines everything downstream.
Four moves turn “student wellbeing” into “Is self-reported sleep quality associated with academic self-efficacy among final-year undergraduates at one UK university?” — a question you can actually answer.
The shortcut most students miss: start from the data
Conventional advice says pick a topic, then find data. Reversing that is often smarter under time pressure. Open a major UK data source, look at what has actually been measured, and let the available variables suggest questions. You arrive at a question you already know you can answer — which is precisely what the first of the four tests demands. Our guide to UK data sources by subject lists the ones an undergraduate can genuinely access.
It is also worth checking your word limit before finalising scope, because a 6,000-word dissertation and a 12,000-word one support very different questions — the verified range across UK departments is set out in our data on undergraduate dissertation word counts. And if your project will be literature-based rather than empirical, decide early which kind of review you are writing, using our guide to literature reviews versus systematic reviews.
Once the question is settled
The topic is the hard part; the blank page after it is the demoralising part. Tesify is built for exactly that gap — you bring your question, your sources and your findings, and it helps you get a structured draft down instead of staring at an empty chapter. It supports honest writing rather than replacing it: the argument, the evidence and the conclusions are 100% written by you, and more than 9,000 students have used it to get moving.
Start your dissertation in Tesify.
Frequently asked questions
Is it acceptable to use a topic from a list like this?
Yes — these are starting points, not finished questions, and every one still needs your population, context and method before it becomes a project. Supervisors are far more interested in whether your question is answerable than in whether the idea was original to you.
Will using an AI tool to help write my dissertation count as academic misconduct?
It depends entirely on your university’s policy, and policies differ across institutions and even between departments. Read yours before you use any tool, declare use where declaration is required, and never submit text as your own that you did not write and cannot explain. Tesify is built to support your writing rather than substitute for it, but the responsibility for complying with your institution’s rules is yours.
What happens to my data and my unpublished work?
Treat this as a question to ask of any tool you use, and check the provider’s current privacy documentation rather than relying on a third-party summary. Your dissertation is unpublished research, and you should know where it is stored and who can access it before you upload it anywhere.
Can I change my topic after it has been approved?
Usually yes, early on, but talk to your supervisor immediately rather than quietly changing direction. A late change can mean re-applying for ethical approval, which takes weeks you may not have.
How specific should my research question be?
Specific enough that someone reading it can tell what data you collected and what analysis you ran. If two people could read your question and picture completely different studies, it is still too broad.
Should I pick a topic I am passionate about or one that is easy?
Pick a feasible topic within an area that interests you. Passion for an infeasible question produces a miserable spring; total indifference to a feasible one produces a competent but dull dissertation. Feasibility is the constraint, interest is the tie-breaker.
What if my supervisor rejects my topic?
Ask which of the four tests it failed. Rejections are almost always about feasibility — data you cannot get, approvals you cannot obtain, scope you cannot cover — and knowing which one lets you fix the question rather than abandon the area.