How to run your side of a supervision meeting

You will get somewhere between eighty and two hundred supervision meetings in your doctorate. That is the entire budget. Most scholars treat these as something that happens to them — a review they turn up to and are assessed in. That framing wastes a great deal of the most valuable resource you have.

Your supervisor is not primarily an assessor. They are an experienced researcher whose judgement you are borrowing while you build your own. The meetings where that works are almost always the ones the scholar shaped.

Send something in advance

The single highest-return habit: send a short note the day before. Three or four bullets — what you did, what you found, what you are stuck on, what you want from this meeting.

This works for reasons that are not obvious. It means the first fifteen minutes are not spent on you narrating and your supervisor catching up; the conversation starts at the interesting part. It gives your supervisor a chance to think before they answer, which produces noticeably better answers than the ones improvised in the room. And writing it forces you to work out what you actually want, which is often the point at which you discover you do not know.

It need not be polished. Four bullets in an email beats a document you did not send because it was not finished.

Bring the problem, not the summary of the problem

The most common way scholars waste supervision time is by presenting work as a finished report when it is not finished, because presenting difficulty feels like admitting failure.

It is not. Your supervisor cannot help with a problem that has been smoothed over. "The analysis is going well, I have a few small issues to sort out" gets you nothing. "I cannot get the model to converge, here is what I have tried, here is the output, I think it is the initialisation but I am not sure" gets you the thing you actually came for.

The fear is that showing confusion reveals you are not good enough. In practice supervisors are usually relieved — a scholar bringing a sharp specific problem is a scholar doing research, and it is far easier to help with than a polished summary that conceals where the work actually is.

Say what you want from the meeting

Start by naming what you need. Supervisors cannot reliably tell which of these you are after, and giving you the wrong one is one of the most common ways a meeting disappoints both of you:

• A decision. You have two options, you cannot choose, you need their judgement.
• A sanity check. You have decided; you want to know if it is obviously wrong.
• Technical help with something specific.
• Context you lack — how the field will react, who else has tried this, whether this is a known dead end.
• Nothing, actually. You are on track and this is a check-in. That is a fine meeting to have, and saying so stops it being padded out.

A scholar who wanted a sanity check and got a forty-minute technical excursion will leave frustrated, and the supervisor will have no idea why.

Write down what was decided, and read it back

Do not rely on your supervisor to keep the record. Some do it well. Many do not, and you are the one who carries the cost of a decision nobody wrote down.

Keep your own notes and, in the last few minutes, read the actions back: so I am running the two clustering approaches, bringing the silhouette scores to the next meeting, and you are emailing the ethics committee about the amendment. This takes thirty seconds.

It catches misunderstandings while they are still cheap. Perhaps a third of the time you will discover you understood something differently — and finding that out now, rather than in a month, is the whole point. It also, quietly, puts your supervisor's own actions on the record, which is the only reliable way they get done.

Disagree, and do it in the meeting

Your supervisor will sometimes be wrong. They will be wrong about your data, which you know better than they do, and occasionally about your field, if your topic has moved past their last close engagement with it.

The failure mode is not arguing. It is nodding, leaving, and then quietly not doing the thing — which reads, from the outside, as unreliability, and means the disagreement never gets resolved. Your supervisor continues to believe you agreed.

Saying "I think that is wrong, because —" is safer than it feels. Most supervisors respond well; the ones who do not are giving you important information about the relationship. Either way you are better off knowing. And the ability to hold a position against a senior person with evidence is not incidental to a doctorate. It is a substantial part of what the qualification is supposed to certify.

When the gaps get long, say so

If meetings keep getting cancelled, or the gap has stretched from a fortnight to two months without anyone deciding it should, raise it.

This is the hardest thing on this list, because it feels like a complaint about someone who holds real power over your progression. But supervisors very often have not noticed. From their side it is two cancellations in a busy term. From yours it is ten weeks without direction.

Keep it factual and forward-looking: we have not managed to meet since February — could we get something in the diary for the next few weeks? No accusation, no apology needed. Most of the time you will get a date, and the problem was never that they did not want to meet you.