my name is natalia. I am a Phd candidate at the radboud university nijmegen medical centre, nijmegen, the netherlands. in daily life, i do research in computational psychiatry. in free time, i go dancing, hiking, investing and blogging.


Recently, I applied for a BKO, a Dutch teaching qualification, as I supervised a few students over the past 1.5 years, and I decided to turn this competence into some more formal achievement. By this occasion, I had to look back into my notes taken during the process and summarise what I learned. Initially, I was hesitating for a long time if I can even manage to supervise a student. Namely, if I will be able to catch some good vibes and keep on going regardless of little bumps on the way which naturally pop out during research: you try and fail, and then try and fail again and again… I was worried that taking a student would pose a lot of pressure on me. And, I felt some responsibility - after all, you have a major impact on these people’s further career choices. Maybe they will get attracted to academia, maybe they will get completely repulsed. But now, I must say that supervising students is a wonderful experience to me, and I consider it – so far – the most fulfilling part of the PhD track. I might be partially due to luck with students, as the ones I got, were extremely motivated and quite self-sufficient. I appreciate the fact that they represent different personalities, because it requires juggling with a variety of supervision strategies from me. 


This experience also made me discover that I have a hidden, second engine. Normally, I program in the morning since in the evenings I am much more prone to errors, and after a few working evenings I am missing energy and lose on productivity. I just can't work effectively for more than eight hours per day. But now, when someone depends on me, I somehow find this additional energy to spend evenings on doing little things for other people's projects just because I know that someone on the other end of the line is waiting for my help. 

And, what I like the most about the whole experience is that the students are just fearless. Once you enter the system and become a PhD student, you might lose this childlike curiosity, and start thinking about other people's expectations too much. Working with students brought me a lot of joy because they were nowhere near being depressed, intimidated or anxious. They just embarked on the project, tried to learn something and do research, without hesitation and without high hopes - and the results are often beyond any expectations.

So, I will always have joyful memories about supervision. Apart from this, there are a few quite functional strategies I developed, which I would like to share:

[a] Results first, literature second. In most cases, students are coming on their very first day as highly motivated and curious about the project… and then, instead of getting a well-defined task, they are getting a bunch of literature to read. A few weeks of reading is enough to make a student frustrated because they have more and more questions piling up while reading, and in the end, the vision of really conducting the project becomes distant, blurry and somewhat threatening. 

In addition to that, others - peers and family - keep on asking: ‘so, what are you actually doing on your project?’… and you need to come up with some general explanation. This is indeed very frustrating. Therefore, I do the opposite: I give the student some exemplary data and some code, and I ask them to produce a first figure with results out of this. They don’t even know what they are actually doing, but they are happy to see some scientific output. I praise them and explain briefly what they actually did - and then, they enthusiastically start reading the literature, and try to figure out what is the meaning of this result, what is the method about and what else they could do in order to improve results further. And in case someone asks them about the project, they always have a result to show. As I examined in practice, this general strategy prevents the student from anxiety and keeps the motivation high throughout the process. 

[b] Good labels. Most people stand on their eyelashes in order to stand up to the standards you put on them. I always comment a student in a positive way from the beginning, namely from the first figure they produce in the first week of their internship. I also introduce them to my colleagues and always with a good word - no wonder that they then start working hard to justify this good opinion. Of course, from time to time it is good to tell the student to redo the task but better to always sandwich this comment between two compliments. 

[c] Details matter. A nice welcome, a farewell, and little gestures really matter to the student. For my students, I prepared a ‘Who is who’ file with a list of photos and short descriptions of all the researchers working in each of the two labs I am working for, so that, once they come, they do not feel like the researchers in these labs are anonymous to them. I think this a little gesture and it took me an hour or two to prepare, but in practice it makes a lot of difference – especially if the labs you are working for do not have their own websites.

[d] Filter the feedback. I myself have a bad previous experience with supervisors which were eager to report my every mistake to their superiors, or were reacting very aggressively when there was a sudden problem like a bug in the code spotted right before a conference or so. Now, I do bot report anything that might sounds pejorative about the student to my superiors, and do not react nervously if there is a sudden problem because I know this will not solve any problem. If I spot a problem, I discuss it with a student instead of complaining to my professors about this. It is very important to stay confidential and only pass the good news upwards, not the bad ones: once you abuse your student’s trust once, the dynamics can go in weird directions, and the student can actually start doing strange and unpredictable moves behind your back. The student must recognise you as a part of their team, as a research partner, and be willing to share all the encountered problems without fearing that this will be passed on to some other person. I always get an alarm bell when the student does not report any problems to me: the problems are always there, so if there is silence on the line, it probably means the lack of trust. 

[e] Let them win sometimes. There was nothing more painful to me at any point since the beginning of my Master studies than encountering people who micromanage. The brain is only able to make a certain number of decisions throughout the day, and pushing a person to spend most of the time on redoing one figure fifty times just because you do not like the colour palette or the font on the axes is extremely frustrating to the student. Therefore, I accept that often, a figure prepared by the student feels like ‘I would plot this differently’ - as long as the fulfils the formal requirements (for instance, for a journal that we are going for), I accept that the student and me might have a different taste, and I don’t waste their time on conforming to mine.

I am still at the beginning of my training process, but I can definitely recommend supervision to any early career researcher. I can tell that now, from a perspective of a supervisor, I also have a better understanding of myself, and a better grasp on mistakes I made as a PhD student. I also better understand my own supervisors’ reactions in certain situations, as they only make sense once you take this new perspective. 

The source of disease

What is the biggest problem of academia?