I’m still rather new to supervising PhD researchers. My first just finished her PhD successfully; right now, I am supervising or co-supervising six more, at different stages and also in different disciplines (blame it on a highly interdisciplinary working environment). The experiences I’ve made as supervisor so far go into my principles together with what I’ve learned from my own PhD project as well as talking to others, seeing what they’re struggling with as supervisors and PhD researchers.

My principles reflect what I am aspiring to as supervisor – they are not to be understood as general recommendations. Some of them may have claims to generality, some are definitely a matter of personal taste and some may be controversial. But decide yourself. Here they come, in a rather random order (though I tried to cluster them thematically):

[Notes: Particularly important principles are highlighted in bold. „They“ means the supervised PhD researchers.]

Be available. Simple and basic as it may sound, I think this is a pretty central one. It doesn’t mean „be available at any time“ – supervision is clearly not your only responsibility. But agreeing to supervise someone’s PhD project has quite some responsibility attached to it. Supervision can only work if the PhD researcher has reliable access to their supervisor(s).

Provide structure, guidance and options, not content and decisions. It’s their project. They need support, but ultimately, it’s their project.

Be open to feedback and criticism. Or, even more strongly: invite them. Good supervision requires self-reflection, and that can only be helped by learning about the perspectives of those you supervise.

Provide feedback and constructive criticism. But gently. Learning to receive criticism is an important part of becoming a scientist. But impostor syndrome is very common in academia across levels of experience. You don’t need to add to it by criticizing in an overly harsh manner. At the same time, this also implies honest feedback. If you think the PhD project isn’t working out, you should communicate that.

À propos impostor syndrome: Let them read The importance of stupidity in scientific research. It can be helpful.

Acknowledge where their expertise exceeds yours. It always does. If not at the beginning of the project, certainly later on. And it just feels well to hear from your supervisor that you know or can do something they don’t.

Let them choose the direction of their PhD project as much as possible. I’ve seen too many overspecified PhD projects with little room for developing own research questions and identifying appropriate methods. Unfortunately, many PhD projects are embedded in third-party funded projects and have to deliver accordingly. But there should be as much freedom as possible – the main result of the PhD project should be an independent researcher. They can’t become one while doing what others have decided they should do. Of course, the supervisor’s job is to support them in focusing on a manageable project.

In other words: Don’t let your preferences dominate theirs. Of course, you’re the experienced one, so you should and must have a say in making choices related to the PhD project. But it should be their choices nonetheless, as it’s their PhD project.

Adapt the mode of supervision to their needs (not the other way around). This may be asking too much from people who have a specific and rigid mode of working. But because the PhD researchers are the ones new to the „business“ and having to learn a lot, I think it’s us supervisors who should be the flexible ones. Some PhD researchers may need close supervision, others prefer to (and can) be left alone. This should influence how supervision is organized on a daily basis.

Communicate your expectations clearly. They should know what you expect and want from them. This also involves finding out whether they feel they can meet those expectations.

Don’t expect too much. They’re just PhD researchers. If you need more, hire a post-doc.

Don’t insist on co-authoring all their papers. It’s become established that the last position in the author list is the supervisor – no matter their actual contribution to a given paper. I believe this is wrong. Supervision of the PhD project should not supplant normal standards of authorship. Unless I have contributed substantially to a given paper, I’m not its co-author – supervisor or not. Having read and commented a draft usually isn’t enough (there’s an acknowledgements section for that).

Allow for distractions and side projects. This shouldn’t of course jeopardize the PhD project, but when one works on one project for three or four years, an occasional intellectual distraction can prevent them from going meshugge.

Put their mental health above the success of their PhD project. Simple as that. It’s only a PhD project. It doesn’t define either you or them. It’s not worth their mental health (or yours, for that matter).

Relatedly: Avoid the sunk cost fallacy (in them and yourself). Some PhD projects simply don’t work out; some people aren’t made for research (which doesn’t mean they’re stupid). In extreme cases, the best option may be to abort the project and move on.

Encourage them to take time off regularly. It’s extremely important in academia to recharge your batteries. Some people need the reminder. Some of those may still not want to take time off as often as you would. But even those should know that it may be a good idea and that it is OK.

Don’t expect them to work more than their contract says. In Germany, part-time PhD contracts are standard. Many supervisors nonetheless expect full-time work. This is wrong. Period. They’re free to work more, but they shouldn’t be expected to. I’m going to die on this hill.

Leverage your network where your own expertise isn’t enough. It’s unlikely your expertise covers everything they need (especially in interdisciplinary contexts). Otherwise, their PhD project wouldn’t be needed. But you’re likely to know other people who can complement the expertise that you can offer yourself.

Provide access to your network. Academia is about knowing the right people who can complement what one can do, so that one can address the research questions one wants to. As supervisor, you usually have an established network that your PhD researchers can benefit from.

Encourage them to exchange with other PhD researchers. These may face similar problems and have solutions or coping strategies only active PhD researchers will know.

Explain how academia works. Academia is a very strange world. To many, it’s a black box. Part of supervision should be introducing them to this strange world and explaining how it works.

Communicate clearly that academia is not for everyone and that doing something else after the PhD is OK. At least in Germany, we can’t offer perspectives to everyone doing their PhD anyways. And there’s a lot of fields in need of smart people with the ability to think scientifically. Ultimately, it’s up to them.

Treat them as colleagues. They’re doing research independently, so they’re not students in the usual sense of the word (though I still think it’s OK to call them „PhD students“, albeit „PhD researchers“ may be preferable), and they’re certainly not your minions.

Note: I’m planning to modify and expand the list in the future. I’m still learning how to be a good supervisor. I will highlight additions and modifications.