I wish I had learned to manage my data sooner, says Matyáš Hiřman in an interview

Who is a data steward and what can the research community expect from one?

Data steward is a new position in the Czech Republic that should help researchers in the long term in the area of working with research data and complying with the conditions required in modern science today. There are many roles of a data steward, however the most important one is to create communication bridges between infrastructures and help smooth running in data management. I personally applied for a data steward course in Vienna in 2022. This was the first pilot course for the position of data steward ever. The purpose of the course, beyond the practical part, was to jointly devise and consolidate the roles and responsibilities of data stewards that arise with the increasing demands of open science and research data management. Until then, research data stewardship responsibilities were mostly "handed to someone," but currently those responsibilities are rapidly increasing. The pilot course was designed to help kick-start active support where it is needed.

Who organized the course for data stewards?

It was at Uni Wien and the course was organised and devised by Tereza Kalová, who coordinates the data stewardship programme at the University of Vienna. The aim was to spread awareness that the position of data steward even exists and should be created across not only Europe but the whole world, so that data scientists are not alone in data management.

What are the specific roles of a data steward?

The role is largely based on where the data steward works. At the Charles University we have three levels of data steward. At the first level, the data steward deals with general data policy issues for the university and also coordinates the data stewards from the different faculties, which is also my position. In the second level, it is the aforementioned data stewards who set the standards in the individual faculties. The third level is the data stewards who are part of the research team. In general, the data steward should help in following research data management procedures.

Do you remember how you first came to the issue of research data management?

I was studying at the Faculty of Science at Charles University, where I am now finishing my PhD, and naturally I had to take care of my research data, which didn't always work as I would have imagined at the moment. I then joined the Central Library of Charles University as a research data methodologist in the natural sciences. When I was looking for where to get my education in data management, my colleagues and I came across a course in Vienna and signed up for it. Gradually it became clear that data steward is a position that is starting to appear regularly in the scientific community and makes sense. It was more of a coincidence at first, but the topic won me over.

How do you hire new data stewards?

This is a complicated question for us as well. The problem is that the position is new, training for it is not yet common, and so the right people are not yet available. The educational opportunities in this field are developing only slowly. Generally speaking, however, it can be anyone who is passionate about data management, or, for example, someone who creates their own programs in their "lab" that make data management easier. We need people who can speak English, who are not afraid to communicate, because their big role is also to educate others. People who are passionate about more technical fields and ideally have a background in the field they will be in. We're counting more on recruiting people that we'll educate along the way.

Are you taking inspiration from anywhere abroad?

The Czech Republic jumped headfirst into data management without any long-term training and education of people. We have contacts with people from the University of Vienna, but we also take inspiration from the universities in Utrecht and Heidelberg. In Vienna, for example, they have five data stewards and a coordinator. That is an example that we would like to have in our environment. We are also part of the 4EU+ (European University Alliance), which is an association of seven universities in Europe where we share our experience.

Is there a special community of data stewards in Europe?

So far only at the level of states, universities and libraries. There is no community of data stewards across Europe, but we still have contacts with people from the Vienna course. However, having an organised community would certainly be interesting.

You have recently become closely involved in the National Repository Platform (NRP) project, what is your role in it?

The project is divided into several key activities and Charles University has taken on key activity seven, focusing on education and human resources. Within this activity, the position of data steward was created, where I have a role. Specifically, the position of data steward science, so I should be looking after topics related to natural sciences. There are also data steward positions for humanities and social sciences and others.

How is this position designed in the NRP project?

We design it as a "train the trainer" position and will primarily respond to the results from each of the key activities of the NRP project. For example, software will be developed to create data management plans, and our role will be to train the people who will use the software, which are specifically the data stewards. We will offer consultations for whole institutions. An important element will be the aforementioned Czech community of data stewards, through which we will test the different parts of the NRP and reflect feedback.

Do you see the EOSC initiative as a bottom-up or rather a top-down activity?

The original idea was to create an environment where communities of disciplines and people who produce and manage data are encouraged to communicate freely and effectively with each other about data management conditions. Unfortunately, EOSC sometimes seems to produce a product that is forced on people. Personally, I don't like centralization and it makes more sense to me to support communities and disciplines, which I am also looking forward to in the Czech EOSC initiative.

How does your work as a data steward blend into your PhD studies?

In retrospect, I wish I'd learned how to manage my data and get it in order earlier. Repeatedly in the sciences we come across the topic of electronic lab journals where data and information can be shared efficiently and easily. This would be extremely important for our data. I regret that I did not know this and no one taught me the various options for effective data management. The active use of my work within my research is probably not very significant as I am (hopefully) in my final year of PhD studies. Researchers often see e.g. data management plans as a bureaucratic burden, which is understandable. But I, on the other hand, think it makes sense in a broader perspective. It's important to know what's going to happen in the project, to plan where and what data I'm going to have and so on, which is something we end up agreeing on even with those who don't like it at first.

What advice would you give to PhD students just starting out about data management?

That would take an hour's lecture. I think it is important to look for possibilities, for example software, to make your work easier and better. Not to be afraid to embrace new things. Just think about data management beforehand, how I name data, where I store it, that it's good to have a plan. The last thing is to not be afraid to go against the flow. PhD students shouldn't be afraid to embrace change or to suggest it directly to their supervisors. They are often the ones who move the scientific community forward.