How to engage researchers with Data Management Plans?
24 May, Esther Plomp & Lena Karvovskaya
On the 11th of April the DCC spring workshop series (a series of Dutch national trainings around research data organised by LCRDM) kicked off with a workshop on ‘How to engage researchers with Data Management Plans’. Fifteen participants gathered to discuss anything related to Data Management Plans (DMPs) and Research Data Management (RDM).
The workshop provided by Esther Plomp (TU Delft) and Lena Karvovskaya (Vrije Universiteit Amsterdam – VU) went into a brief introduction of how DMP and RDM support looks like at their universities. At TU Delft, Data Stewards are responsible for providing feedback (within 1-2 weeks) on DMPs. Researchers may use (Word) templates and are encouraged to use the online platform DMPonline. For projects with privacy/ethic concerns, the TU Delft template on DMPonline is required. At the VU, Data Stewards or RDM specialists at the Library are responsible for feedback. Researchers are encouraged to use the online platform DMPonline and the VU template but may use (Word) templates. At both TU Delft and the VU DMPonline is used as a GDPR registry (keeping track of projects that process personal data across the organisation, in accordance with European legislation).
Both Esther and Lena also showed examples of how to track incoming requests from researchers on DMPs and other research support requests. Esther keeps track using an Excel spreadsheet that covers information such as the type of request, department of the researcher and the duration of the request. Lena showed how many requests come in at the RDM Support Desk at VU Library; these numbers don’t show the work of faculty data stewards. Tracking your incoming support requests will allow you to better estimate how long requests will take and provide you with an overview of where you spent your time on. This will prove helpful in annual reviews or other times where you have to account for where your time goes. Solutions that participants of the workshop used to track support requests included: Jira, Azure Boards, inhouse software solution, and ticketing systems like Service Now.
Lena discussed that in order to engage researchers with DMPs, it is helpful to take the researcher’s perspective. This will make it easier to understand why there are many reasons that even researchers who are aware of existing support usually do not to reach out to an RDM specialist for support. Researchers are busy individuals (see an overview of tasks that are only associated with the RDM part of research!). Often, researchers only reach out to support when things are going wrong and there is a clear reason to contact support. Researchers may perceive many barriers to contacting RDM support – such as that their question is not big enough or could be stupid. Researchers often perceive Research Data Management as a number of bureaucratic tasks that need to be fulfilled before actual research can start. We should therefore see setting up the DMP as an opportunity to engage with the researcher which may improve their RDM practices before things go wrong and data is lost. DMPs also offers a unique opportunity to reflect on data management in a holistic way, to show it as integral part during all steps of the research process.
After this short plenary introduction on DMPs and RDM requests, it was time for an exercise! The first exercise took about 45 minutes of the workshop and focussed on reviewing a DMP and discussing feedback with fellow participants. The participants had 15 minutes to review a DMP (with a choice of several public DMPs on topics such as robot software, migration data, isotope archaeology, Covid and climate). After reviewing a DMP individually, participants had about 15 minutes to review each other’s comments and discuss the feedback in pairs or groups. After these smaller discussions we had 10 minutes to discuss how the exercise went with the entire group.
In this plenary discussion, we agreed that DMPs are more user-friendly when they are short, specific, clear, responsibilities are outlined, and when any external references include working links. This was not always the case in the DMPs that were reviewed in the exercise: these DMPs were written for large-scale projects in an early stage. The DMPs often did not clarify which solution researchers would use, and instead mentioned a lot of potential solutions that are available. The reason for this lack of specificity might be that the DMP is written only by one partner within the large collaboration and that there is no consensus yet between stakeholders as to what solutions to use. Esther highlighted that DMPs are an excellent opportunity to reach this consensus on what solutions to use. If there is an opportunity for the data steward to attend the project kickoff meeting, the data steward can refer to the gaps in the DMP to reach consensus and ensure that the DMP will be helpful and applicable to the project.
We discussed that it can be helpful to see DMPs as part of project deliverables (like the European Commission does, which provided us with the publicly available DMPs used for the exercise!). We had open questions about whether it would be a good idea to update the plan when the project is already finished. We ended on the question whether estimates of data storage requirements are sufficient – which we thought it was, as long as the estimates provided enough information to connect researchers to the appropriate storage solutions (for example: don’t point researchers to a storage solution of less than 1 TB if the researchers estimate that they would generate more than this).
In the second part of the workshop we focused on ideas that would make researcher’s lives easier. Esther highlighted the possibilities of providing researchers with templates via Word or DMPonline, setting up information on a support website, or sending out newsletters (keeping in mind that people may not read these :)). A more advanced way for researchers to structure the workflows in their labs are Team Manuals or DMP statements. Both manuals or statements can describe expectations for team members, highlight resources and provide documentation on processes, which has the potential to make research workflows more efficient. After these examples it was again time for another exercise that lasted about 40 minutes on sharing ideas or existing practices to make researcher’s life easier.
Ideas that the workshop participants shared were:
- Have a central location with information about RDM/DMPs (such as intranet)
- Have a one-stop-shop contact for consultation (for example, a ticket system), or make it clear who is the dedicated person for what topic, so that questions can easily be directed and answered.
- Minimalisation of different forms/workflows – avoid overlap!
- Drop-in days/hours
- Trainings (also online!) and e-learnings
- When there is a meeting taking place try to get in and explain what a data steward can do – stay after the meeting to answer more questions!
- Try to be part of a project (although this is not possible for every data steward)
- Provided clear and informative sample answers in DMP templates
- Use example DMPs from other researchers (with permission!) or connect researchers with similar projects with each other.
- Provide regular reminders for DMP updating – either a tool or discuss this regularly during meetings
- Provide a timeline to clarify expectations (for example, the 1-2 weeks for feedback on a DMP).
- Consistent feedback – talk to colleagues about their feedback!
- Work with the communication department – it is not only offering the tools and training, it is also about communicating this!
- Do RDM onboarding – talk to starting people or join obligatory courses/onboarding meetings
- Don’t over-communicate or provide too much information.
- Have a pitch about what bad things can happen if you don’t have a DMP.
- Go to informal coffee meetings! This will lower the barrier for people to approach you and find out about low-key problems before they become big
- Candy helps!
We concluded that not all of these solutions will always be applicable within your own local context. Different scales require different approaches. Are you supporting 1000 or 10 people? Are you located centrally or decentralised? Whether your initiatives will be successful or not will also very much depend on whether you are effectively communicating them! We don’t want to spam researchers and ensure that they receive useful information at the right moment. As data stewards we usually don’t have training in marketing and communication, but we can ask our colleagues from marketing and communication for help.
Feedback from the participants (gathered via a google form) on the workshop was overly positive (giving the workshop an 8,2!). Particularly the interactivity of the session was appreciated, as well as the opportunity to network. Points for improvement include forwarding the workshop agenda in advance, also doing introductions of the participants, include more up-front information on how other institutes handle DMP/RDM requests, and focusing more on how to practically engage researchers with DMPs (by for example providing more information on the benefits of DMPs for the researcher).
We hope that these ideas and suggestions are helpful to others who were not able to participate in the workshop – and hope you may still sign up for the next DCC spring courses via the sign-up form! See you there!
Open Working 