JISCMRD Workshop: "Meeting (Disciplinary) Challenges in Research Data Management Planning"

As our project came officially to an end on 31 March (although we will continue until the end of May on an unfunded extension), we participated, together with the other projects in our strand, to the JISCMRD Workshop: "Meeting (Disciplinary) Challenges in Research Data Management Planning". The workshop was designed to allow projects to share their final findings and experiences with other projects and with JISC. The workshop, which took place on 23 March at Etc Venues Paddington, has been a very good occasion to better understand what similar projects have been working on in the past 6 months. Although projects shared their developments on their respective blogs, listening to a summary presentation of the results has been extremely useful to identify similarities between other projects and ours that we might have previously missed. In fact, I realised that some of the observations I made in a previous post on our approach to RDM also apply to other projects as well.

The first part of the workshop was dedicated to the presentations by all the projects. I will not summarise them all (see here), but only point out similarities with the SMDMRD project, and what we found most useful and interesting.

The goal of the DATUM in Action project (Northumbria University) was to support research staff on a EU project to plan and implement research data management. Their approach has been similar to ours: collect and analyse requirements, create a data management plan, and provide the supporting technical infrastructure. In common we also had a series of complex requirements, such as for example the need for data version control. Their approach has been different in the sense that they decided to use a combination of different tools (CMS, distributed version control, shared drives), some proprietary (MS Sharepoint), and others OpenSource (BZR).

The DMPS for Psychology project won the prize for the most reusable results, and we definitely agree with the vote. We will certainly use their website as an example to follow for our online material on research data management, both at the research group level and at the institutional level.

Interesting results were also pointed out by the History DMP project (University of Hull), such as the DMPOnline "pre-filled" with information from the institutional/departmental data management policies. The project's team had the advantage of having previous experience in the subject, and a well-established repository (Hydra). Some of Hydra's features (Fedora, linked data, DataCite, the very slick look) make it very interesting for us in a longer term perspective.

With the Research Data Management for Mechanical Engineering Departments (REDm-MED) we definitely have in common the large number of different file formats and standards, and the difficulty to make people add descriptive metadata to the data sets (especially regarding the relationships among different files in a set) in order to make them reusable in the future, although we might not follow their approach based on RAID (Research Activity Information Development) diagrams.

Finally, two concepts pointed out by the REWARD project were of particular interest for us: the proposal of letting PhD students create a data management plan (using for example the DMPOnline); and the recommendation to use discipline-specific repositories (e.g. a data journal, or a repository across institutions), and sharing only metadata on the respective institutional repositories, something we suggested to do with our local DSpace installation and the Queen Mary Research Online repository (see for example this post).

The second part of the workshop was dedicated to the DMPOnline tool. A new, improved version was presented. Of the new version we appreciated the extension to additional funders and the customisation features. We were also asked what features we would like to see implemented: among them, the ability to share and collaborate in the creation of plans would be very important. One thing we agreed with others is the strong focus on "funded" projects (the very first question asked is "Budget"): even tough the tool can be used for any project, the mere question creates a barrier to its adoption, for example by PhD students or research groups like ours (see the previous post on creating data management policies). Finally, very insightful was the presentation by David Shotton who analysed and compared tools for creating Data Management Plans online (see his blog for details).

To conclude, the workshop was very useful to broaden our still quite limited view of the subject, and to get a complete update on the results of the various projects (unfortunately it is always hard to find the time to regularly follow all the blogs).