Supporting researchers with their data management plans

What does the DMP support look like in your institution? How do you deal with discipline-specific issues related to data management? What is your definition of “data”? These are just a few topics that were discussed during the first knowledge hub lunch session in May 2021. With about 30 participants, most of them data stewards/librarians and research support staff, we had a lot to talk about.

We kicked off with flash presentations by RDM stewards from KU Leuven, UGent and UHasselt who talked about their approach to DMP support. It soon became clear that we were all dealing with the same challenges. Researchers are overwhelmed by (the administrative effort of) a DMP and do not necessarily speak the same “data” language as we do. It was inspiring to see the solutions that the three institutions came up with to overcome these issues, from elaborate information on their website to training sessions on a specific difficult DMP category to one-on-one talks with researchers. 

The importance of taking time to talk with researchers was expressed multiple times in the discussion afterwards. Do researchers believe that they have no data? Let them tell you what their research is about and you will discover the data within. Do you not understand the data because you are not familiar with the discipline? Just ask them. Researchers will gladly explain to you what kind of data they work with, how these are managed and what they need help with! By listening, we can remove the “language barrier” between researchers and data stewards.

Another burning question that came up was whether you should provide researchers with model answers for certain sections of their DMP. On the one hand, model answers and examples could help researchers understand what is being asked of them and what would be seen as best practice. Some sections of the DMP are indeed so standard that it would not be an issue. On the other hand, it could stop researchers from really thinking about what management is actually best for the data they generate. And the data management procedures proposed in the plan may not actually be implemented.

Follow-up of actual implementation of DMPs was also discussed, acknowledging that this is not always easy. Whilst it is the responsibility of researchers, there are ways in which support staff can help with this. Having a data manager within a research group, working in close contact with data stewards, was mentioned as helpful. And monitoring the publishing or sharing of research data at the end of a project can be a way to see how well a DMP was implemented. 

If we conclude that this first session was a huge success, that statement is backed up by almost all participants indicating that it was (more than) satisfactory. Most appreciated was the opportunity to share (often similar) experiences and learn from the approaches, ideas and practical examples of colleagues. We look forward to a live discussion sometime soon, but for now end with the wise words: “imagine your PC crashing … What have you lost? Those are your data!”