For an examined qualification, your supervisor must normally be a member of staff at your school, and that is a rule you cannot work around — an outside expert can advise but cannot replace them. For independent research, anyone competent can supervise, including a PhD student, and in practice a PhD student is often a better supervisor than a professor because they have the time. What matters is not the title but whether the person reads your drafts.
"Who can supervise my research?" has two different answers depending on whether your project is examined.
Get this distinction wrong and you can invalidate a qualification, or spend months looking for permission you never needed.
If your project is examined, the answer is narrow
For an EPQ, an IB extended essay, a Scottish Advanced Higher project or AP Research, the supervisor of record must be a member of staff at your school or college. This is not a preference. The supervising role carries assessment and safeguarding duties that only the centre can discharge.
AQA's guidance on EPQ safeguarding requirements gives a good sense of what that actually involves. The centre must engage with students about topic choice and steer them away from topics that could harm them; must refuse topics grounded in personal experience of things like eating disorders, exploitation or abuse; must check that the topic and any primary research suit the student's age and maturity; must consult the Designated Safeguarding Lead before approving a title; must complete a risk assessment where primary research involves vulnerable people or sensitive subjects; and must document safeguarding measures where a student discloses sensitive information. AQA states that failing to follow the relevant JCQ regulations "is considered centre staff malpractice."
No outside academic can do any of that. They have no relationship with your school's safeguarding lead, no access to your records, and no standing with the exam board.
**What an external expert can do** is advise on substance. You can email an academic about your methodology, incorporate what they say, and acknowledge it. Your school supervisor remains the supervisor. If an external mentor is involved substantially, tell your school supervisor — most boards require declared assistance to be recorded, and an undeclared mentor is the kind of thing that causes real problems late.
Our comparison of the EPQ and the extended essay covers how the two qualifications differ on supervision, and our complete guide to the EPQ goes through the supervision log in detail.
If your project is independent, the answer is broad
For research you are doing outside a qualification — to publish, to enter a competition, or because you want to — there are no rules about who supervises you. There is no body to satisfy. The only question is whether the person is competent and available.
Here is the hierarchy that actually matters, which is close to the inverse of the status hierarchy:
| Who | Subject knowledge | Time available | Usually best for |
|---|---|---|---|
| PhD student in your area | Current and narrow | Often good | Almost everything |
| Postdoc | Current and broad | Moderate | Method design |
| Lecturer / assistant professor | Broad | Limited | Framing, venue choice |
| Full professor with a group | Broad, sometimes dated at the edges | Very limited | A letter, a door |
| Subject teacher at school | Variable | Good | Structure, writing, feasibility |
The counter-intuitive entry is the top one. A PhD student is doing original research right now, reads the literature in their narrow area weekly, and remembers being confused by exactly the things confusing you. They also have time in a way that a professor running a group does not.
The most common mistake is to chase seniority. A professor's name attached to a project they did not read is worth less than a PhD student's sustained attention, and it is obvious to anyone reading the work which one you had.
Why academics so rarely say yes
It is worth understanding the constraints, because they explain why a reasonable request goes unanswered.
The arithmetic is brutal. The Goldhaber-Gordon lab at Stanford publishes an unusually frank page about research experiences for high school students. It states that the lab receives "tens of such requests per year" and takes a high school intern roughly once every four to five years, because its own PhD students, postdocs and undergraduates take precedence. That is not a hostile lab — it is a lab that wrote down what everyone else leaves implicit.
Many labs legally cannot. The University of North Carolina's minors in laboratories policy sets out the age thresholds, training and supervision requirements that apply to under-18s in a research laboratory. Comparable rules exist across US and UK institutions. A wet-lab group saying no may have no discretion at all.
Volume response rates are low. The University of Arizona's careers service advises contacting three to five faculty members rather than one, which tells you what the base rate looks like even for well-written requests.
The practical implication: computational, archival, data-based and theoretical groups face neither the safeguarding constraint nor the physical-access constraint, so start there. Our guides to research without a lab and remote research opportunities cover what is possible.
How to ask so that someone might say yes
The Stanford page is also useful as a template, because it says what the lab wants to see in an email: why you are writing to that specific person, demonstrating "substantial familiarity with their work"; what you bring — skills, grades, sustained achievements; the time commitment you are offering, in hours per week and over what period; and your goals for the experience.
It also makes two recommendations worth repeating. Prefer local mentoring — remote supervision is "harder on both sides," particularly for experimental work. And focus on fewer, well-researched leads rather than mass emailing, noting that one personalised email may take one to two hours to research and write.
A fourth suggestion from that page is the most under-used advice in this whole area: consider smaller institutions and national labs, which may have less access to students and can therefore offer more attention. Everyone emails Oxford, Cambridge, MIT and Stanford. Far fewer email a research-active department at a less famous university where a PhD student has capacity and would enjoy the conversation.
Our guide to emailing a professor about research covers the message itself, and how to find a research mentor covers where to look.
The test for whether supervision is real
Whether your supervisor is a teacher, a PhD student or someone you found yourself, one question separates genuine supervision from a nominal arrangement:
Will they read a full draft and write on it?
Not meet to discuss it. Not take a call. Read the whole thing and return it marked. Written feedback on a complete draft is the single intervention that improves student research most, and it is the one that costs a supervisor real time — which is exactly why it is the honest test.
Two follow-up questions worth asking before you commit:
- What are you working on at the moment? A specific answer means current engagement with the field. A vague one means you will be getting general advice.
- How many drafts will you look at? One is a courtesy. Three is supervision.
If you are evaluating a paid programme, apply the same test to it. The question is not who is on the website; it is who reads your work, and how often. Our guides to what a research mentorship programme is and research mentors explained set out what good supervision looks like in practice.
In short
For anything examined: school staff, no exceptions, with external input declared. For anything independent: whoever is competent and will actually read your drafts — and a PhD student in the right subfield is usually a better bet than a professor in a famous one.
If nobody says yes, you can still do the work. Our guide to independent research in high school covers doing it unsupervised, and mentored versus self-directed research sets out honestly what you gain and lose either way.
Frequently asked questions
Can a university professor supervise my EPQ or extended essay?
Not as the official supervisor. For an examined project, the supervising role sits with your school, which has assessment and safeguarding responsibilities an external person cannot discharge. An academic can act as an additional subject contact, and you can cite their input, but your school supervisor remains the supervisor of record.
Is a PhD student good enough to supervise research?
Often better than a professor. A PhD student is doing original research now, knows the current literature in their narrow area, and — crucially — has time to read drafts. A professor with a research group, teaching load and grant deadlines usually does not. For school-level research, availability beats seniority.
Do I need a supervisor to do independent research?
No. Nothing prevents you doing and publishing research alone, and student journals accept single-author submissions. A supervisor shortens the process considerably by catching design problems before you collect data, which is the stage where mistakes are expensive.
Can my parent supervise my research project?
For independent work, if they have genuine expertise in the area, they can give useful feedback — but declare the relationship if you submit anywhere, because undeclared family involvement reads badly. For an examined qualification, no: the supervisor must be school staff, and a conflict of interest would be a serious problem.
How do I know if a research mentor is any good?
Ask what they are currently working on and whether they will read full drafts. A mentor who can describe their own live research in specific terms, and who commits to reading your work rather than meeting to discuss it, is doing the job. One who offers calls but not written feedback is not.
Why do so few academics take on school students?
Time, and legal constraints. One Stanford lab states publicly that it receives tens of requests a year and takes a high school intern roughly once every four to five years. Separately, many laboratories cannot host under-18s at all because of minors-in-labs policies. Neither is personal.