A second-year doctoral student put it like this: “My supervisor is brilliant and well known, but I saw him twice last semester, fifteen minutes each time. A draft I sent three months ago still has no comments.” A friend in another group, supervised by someone younger with fewer publications, meets weekly, gets feedback within ten days and already has a first paper. A great researcher and a good supervisor for you are not the same thing.
In many programmes, especially research-only models, choosing a supervisor is the single most important decision of the entire PhD. It shapes the direction of your project, the resources you can draw on, and whether the next three or four years feel manageable.
Three layers of fit
- Intellectual fit: are they working now, not just in the past, in an area close to your project? Look at their last five years of publications, not only their famous older papers.
- Supervision style fit: some supervisors set detailed tasks; others give almost complete independence. What do you need?
- Practical fit: do they have funding, time, and plans to stay at the institution for the next few years?
Most applicants only check the first layer. The other two cause most of the trouble.
Supervision styles: no single right answer
| Style | What it looks like | Suits | Risk |
|---|---|---|---|
| Hands-on | Frequent meetings, specific milestones, detailed edits | People new to research who want structure | Less room to build independence |
| Hands-off | Meets when needed, candidate leads | Experienced, self-disciplined researchers | Drifting off course without anyone noticing early |
| Large-group | Day-to-day guidance via postdocs and lab meetings | People who enjoy busy teams and peer learning | Limited direct time with the principal supervisor |
| Project-based | Topic tied to an existing funded project | People wanting a clear topic with resources | Less freedom to choose direction |
The first email
Academics receive many supervision requests, most of them generic mass emails. The ones that get replies usually:
- Are short: one phone screen.
- Show you have read their work: mention a specific paper and how it connects to your idea, not generic praise.
- State your research idea in two or three sentences: the question, why it matters, a rough approach.
- Summarise relevant background: your master's thesis, methods skills, practical experience.
- Ask one concrete question: are they taking students next year, could you have a twenty-minute call?
Attach a short CV. Do not attach a twenty-page proposal to a first email.
Questions to ask in the conversation
- “How often do you usually meet your doctoral students, and in person or online?”
- “How long does it typically take you to comment on a chapter draft?”
- “How do your students usually arrive at their topics: their own proposals or group projects?”
- “What are your expectations about publishing during the PhD?”
- “How long have your recent students taken to finish, and where are they now?”
- “Do you have plans for extended leave or a move in the next few years?”
The last one feels awkward, but supervisors moving institutions mid-project is a real and disruptive scenario. You can frame it gently: “I'd like to picture how the group will look over the next three or four years.”
Talk to current and former students
This is the most valuable and most neglected source of information. Ask the supervisor for contacts or find them on the group page. Speak privately, without the supervisor present, and ask open questions:
- “How does your supervisor react when experiments or drafts go badly?”
- “Did you feel supported when you had personal difficulties?”
- “What is the group culture like: hours, weekends, collaborative or competitive?”
Notice what is not said. Hesitation, a flat “it's fine”, or “you should ask others too” are worth thinking about.
Warning signs
- Several students have left or switched supervisors in recent years.
- They cannot answer simple questions about meeting frequency or feedback times.
- They overpromise at the first meeting: guaranteed top-journal papers, early completion.
- They criticise colleagues or former students in front of you.
- The group has far more students than the supervisor could realistically support.
One sign alone means little. Several together deserve serious thought.
Co-supervision and supervisory panels
Two supervisors, one strong in the subject and one strong in methods, or one senior and one early-career with more time, can be an excellent arrangement. It also lowers the risk if one becomes unavailable. Settle three things early: who holds primary responsibility, in what order drafts go to each, and how you handle conflicting advice.
Once you have chosen: agree expectations early
After you are accepted, suggest a meeting to agree expectations and record them in a short document: meeting frequency, how drafts are sent, expected turnaround for feedback, working-hours expectations, and how authorship on papers will be handled. Many universities provide a supervision agreement template; if not, a summary email after the meeting does the job.
Start today: list five potential supervisors, read the two most recent papers by each, and write one sentence on how their work connects to your idea. If you cannot write that sentence for someone, they should not be on the list yet.
Câu hỏi thường gặp
How do I choose the right PhD supervisor?
Check three layers: whether their current work is close to your topic, whether their supervision style suits you, and whether they have funding, time and plans to stay. Talking to their current students is the best way to verify.
How should I write an email to a potential PhD supervisor?
Keep it short, mention a specific piece of their work that connects to your idea, state your research question in two or three sentences, summarise your background and ask whether they are taking students. Attach a short CV, not a long proposal.
What questions should I ask a potential supervisor?
Ask how often they meet students, how quickly they give feedback on drafts, how topics are chosen, what they expect about publishing, how long recent students took to finish and whether they plan to move.
Is it a good idea to choose a famous professor as a supervisor?
A big name can bring networks and resources, but if they are too busy to meet and give feedback, the PhD can be hard going. Time and supervision style matter as much as reputation.
Is having two PhD supervisors a good idea?
Often yes, because it combines expertise and reduces risk if one is unavailable. Agree early on who has primary responsibility and how conflicting feedback will be handled.
