PhD Supervisor Red Flags

Treat each item below as a question to answer, not a verdict. One signal is information. Two or three that point the same way, or one that survives a direct question, is a reason to choose someone else. Three rules keep the exercise useful.

Check before you conclude. A group page that has not been updated since 2023 might be a group in decline or an administrator who left.

Ask the person directly. How a supervisor answers an awkward question is itself the most informative thing you will get, and a good one will not be offended by it.

Ask students, not only the supervisor. Current and former students answer more honestly than any page, and being refused access to them is a stronger signal than anything they could have told you.

The publication record is public, and it says more about how a group works than the group page does. Look at the last three to four years only, and look at authorship positions rather than counts.

No new first authors. If every recent paper is first-authored by the supervisor or by the same one postdoc, students in that group may not be leading work of their own. Check the first-author names against the group's people page and its alumni list.

A long gap in output. Nothing published since 2022 or 2023 can mean a grant ended, the group wound down, or the supervisor moved into administration. Check whether recent papers exist under a new affiliation before concluding anything.

Students appearing once and never again. Look up two or three recent first authors and see whether they published a second and third paper from that group. A single paper followed by silence sometimes means someone left.

Everything published with the same external collaborator and nothing without them. That can mean the intellectual centre of the project sits elsewhere, which changes who you would actually be trained by.

Retractions or corrections concentrated in the group. Individual corrections are ordinary and mean nothing on their own. A pattern is worth understanding before you attach four years to it.

Turnover is the thing applicants most want to know and most struggle to see, since nobody advertises departures. It is still checkable with two searches and one question.

An alumni list that does not exist, or that stops several years ago. Groups that keep students usually enjoy listing where they went.

People disappearing between versions of the people page. An archived copy of the page from two years ago, compared with today's, shows you who left. Names that appear and vanish inside eighteen months are worth asking about.

Doctorates that do not appear in the university's thesis repository. Most institutions deposit completed theses publicly, so a group with many students over ten years and few deposited theses is a question.

Nobody available to talk to you. A supervisor who will not put you in touch with a single current student, or a student who answers only with the supervisor copied in, has told you something.

The direct version of the question, which is entirely reasonable to ask: how many students have you supervised, how many completed, and where did they go?

Vagueness about funding is the most common and the most costly warning sign, because it is easy to hear as optimism. Funding questions have precise answers, and someone who has the money can give them in one line.

"We will find something" as an answer to a direct question about funding. Ask which scheme, which deadline, and what happens if it is unsuccessful.

Funding that ends before the thesis is expected to. If the project is designed as four years and the grant runs three, ask what previous students did in the final year.

An invitation to self-fund a project the supervisor designed and wants done. That is a request to pay for your own labour on someone else's research agenda.

No named grant or scheme anywhere

not in the vacancy text, not in recent paper acknowledgements, not in the answer to your question. Grant numbers are public and supervisors quote them easily.

Pressure to accept before you have the funding details in writing. A real offer survives a request to see the terms.

The correspondence before you commit is the only sample of the supervision you get in advance. It is a small sample, but it is not nothing, and behaviour that appears while they are recruiting you rarely improves afterwards.

No interest in your ideas. If two exchanges have gone by and nobody has asked what you want to work on, you are being recruited as hands rather than as a researcher.

Unpaid work as a test

an analysis, a literature review or a written piece requested before any offer exists. A short technical conversation is normal, weeks of unpaid output are not.

An offer made within hours of a first contact, with pressure to accept. Genuine recruitment involves a call, questions in both directions and a delay.

Authorship offered as an inducement, especially first authorship on work that already exists. Authorship follows contribution and is agreed after it, and treating it as currency in a recruitment conversation is a preview of how it will be handled later.

Contradictions between what you are told and what the department's pages say, about duration, teaching load or the funding route. Ask about the difference and see how it is handled.

Over-reading warning signs costs applicants good positions, and several of the things that worry people most are ordinary. None of the following is a reason to walk away.

Being busy and slow to reply. Two weeks of silence on a cold email is the norm and says nothing about the person.

Being demanding about the work. High standards, hard questions and rewritten drafts are what supervision is for. What matters is whether the criticism is about the work rather than about you.

A small group or an early-career supervisor. New group leaders often give more attention, have fresh funding and are strongly motivated to see their first students succeed. The trade-off is a smaller network and less accumulated experience, not worse supervision.

Not having a place this year. Capacity is finite and varies annually. It is a scheduling fact rather than a judgement.

Direct rejection. A supervisor who tells you plainly that your profile does not fit has done you a favour compared with one who lets an application drift.

A brief, telegraphic reply to your first email. Short is not cold, and many productive exchanges start with three lines.