How to Choose a Psychology Dissertation Topic in the UK (2026)

Short answer: Two things decide a UK psychology dissertation before the idea does: which population you can lawfully recruit, and what recruitment will cost. Clinical and patient populations run through NHS research governance, which undergraduates cannot enter and most taught postgraduates cannot either. Online recruitment is open to everyone but is not free. Settle both questions in week one, because they determine whether the title you want is a title you can earn.

The two BPS documents your ethics form will be judged against

UK psychology ethics committees work from the British Psychological Society’s Code of Ethics and Conduct, current version published December 2021, and the Code of Human Research Ethics, current version April 2021. Neither has been superseded. Cite them by year rather than by edition number, because the BPS does not print edition numbers on the current versions.

Read the Code of Human Research Ethics before you write your proposal, not after your application is returned. The four principles it turns on, respect, competence, responsibility and integrity, are the categories your reviewers will use, and writing your rationale in those terms shortens the review considerably.

Why your clinical topic becomes a student sample

This is the single most common trajectory of a UK psychology dissertation. The proposal says “anxiety in patients referred to IAPT”. The ethics committee explains that NHS patients require Health Research Authority approval. The HRA does not accept applications from undergraduate students at all, and restricts master’s eligibility to students on health or care programmes in research-active departments. The project becomes “anxiety symptoms in a university sample”, and the student writes a discussion that quietly generalises to a clinical population it never touched.

There are two honest ways out, and both are better than the drift.

• Design for a non-clinical population from the start and make the question genuinely about that population. Subclinical variation in a general sample is a legitimate object of study, not a consolation prize, as long as the title and the discussion say so.
• Use existing data. The UK Data Service holds major studies at Open and Safeguarded access tiers that a registered student can use. Secondary analysis of a large, well-documented dataset will usually produce a stronger dissertation than an underpowered original study, and it removes recruitment entirely.

What recruitment actually costs

Prolific is the realistic default for UK online recruitment. Budget honestly before you finalise the design. Prolific’s recommended participant rate is £9.00 an hour with an absolute minimum of £6.00, and the academic platform fee is 33.3% on top of what you pay participants.

So a 2×2 between-subjects design needing roughly 200 participants for a 15-minute study costs about 200 × £2.25 in participant payments, plus a third again in fees: comfortably over £500. Most departments do not fund this. Find out your department’s position on participant payment before you commit, because the answer often decides between an experiment and a survey.

Run a power analysis at the design stage rather than reporting one afterwards. An underpowered study with a null result is not a finding, and examiners know the difference.

Preregistration is now expected, and it is quick

UK psychology departments increasingly expect a preregistration, and it protects you: a preregistered null result reads as a competent test rather than a failed project. AsPredicted asks nine questions and takes about twenty minutes. The Open Science Framework offers fuller templates and is free.

Preregistration is also a viable dissertation topic in itself. OSF’s registry metadata is public, so adoption rates, discrepancies between preregistered and reported analyses, and the growth of Registered Reports can all be studied without recruiting a single participant.

Where the material is

Live territory for 2026-27

• Online Safety Act implementation. Ofcom’s Protection of Children Codes came into force on 25 July 2025, requiring highly effective age assurance and annual children’s risk assessments. Adolescent wellbeing, the friction age assurance introduces, and displacement to unregulated platforms are all open empirical questions, and can be studied with adults reflecting on their own use where recruiting minors is not possible.
• Parasocial and therapeutic use of AI chatbots. Fully researchable through an online experiment or survey on a general adult sample. No clinical population required, and very little settled literature.
• Replication and research practice as an object rather than a method, using public registry data.

If your topic touches self-harm, suicide, eating disorders or trauma, expect a substantially longer ethics review, a requirement for distress protocols and signposting, and in many departments a rule that undergraduates may not run it at all. That is not a reason to avoid the field, but it is a reason to confirm the position with your supervisor before you write the proposal.

Sources

• BPS, Code of Ethics and Conduct: https://www.bps.org.uk/guideline/code-ethics-and-conduct
• BPS, Code of Human Research Ethics: https://www.bps.org.uk/guideline/bps-code-human-research-ethics
• Health Research Authority, student research eligibility: https://www.hra.nhs.uk/planning-and-improving-research/research-planning/student-research/
• Prolific: https://www.prolific.com
• UK Data Service: https://ukdataservice.ac.uk
• Open Science Framework registries: https://osf.io/registries
• AsPredicted: https://aspredicted.org
• Ofcom, Protection of Children Codes: https://www.ofcom.org.uk