DMP discussion areas

During June 2017 an interesting discussion arose on the researchdataman jiscmail list. We thought a digest of responses would be a useful addition here.

DMPs for PGR students

Mary Donaldson (Glasgow University) commented, *"I'm currently looking into how we handle PGR data at Glasgow, from our policy expectations, to what actually happens, to what would happen if all our PGR students decided to comply fully with out policies tomorrow.

Something that has been floated occasionally is making a data management plan part of the documentation that is required for annual progression. This is something that some grad schools are keener on than others. There would of course be an exemption for students in disciplines which don't have data.

I was wondering if anywhere already does this, or anything along similar lines? If so, what were the justifications used for introducing it?"*

There are a number of strategies currently being used. James Davenport (Bath) stated that Bath is making a DMP part of 'confirmation of candidature, i.e. the 1-year progress report. "There is no per-discipline exemption as such, though the DMP could be 'we don't expect to produce any data'.
I personally would oppose per-discipline exemptions, as who knows what an individual student may do. A classicist may have a database of different readings, for example, or a pure mathematician computational experiments."

Kevin Ashley (Director, Digital Curation Centre) commented that at least two UK universities he knows of (Bath and University of East London) have addressed PGR data through the provision of customised guidance on data management & the use of simplified templates for their plans. In both cases the justifications were two-fold:

• PGRs need to have a plan because of broad university policy that applies
  to all research (that is, you would need a justification to exclude them
  from that requirement
• Given that, a more lightweight template can be justified to make the process
  easier for them and those in the university who need to review them.

LSHTM takes a similar approach, Gareth Knight reported, "We have an 'expectation' that MPhil/PhD/DrPH students should write a DMP for inclusion in their Upgrading / Review report. The feeling at the time was that it should be recommended for the first few years so that lecturers could become accustomed to the new document, and it would be mandated at a later date. The only exemptions we have are for consultancies and projects using secondary material only. Nothing for specific research departments/groups."

Laurence Horton (LSE Library) stated that LSE requires DMPs as part of research ethics review. *"Ethical review is mandatory for any research involving human participants, or data relating to identifiable human subjects. Of course, that means project that don't involve human subjects, or get ethical approval at other institutions or bodies (NHS for example) don't have to submit a DMP so there's a lot that fall through the gap. Also, if researchers want to apply for sensitive data the library holds or manages access to, for example Eurostat or UK Data Service special licence, we will require a data management plan before we submit their request.

We have a customised research staff template in DMPonline for non-RCUK DMPs.

This obviously covers all research including PGRs. It's not part of formal part of academic progress.*".

Georgina Parsons (Open Access & Research Data Manager) commented that Cranfield made DMPs mandatory for doctoral students in October 2016. *"Our justifications were to meet our own RDM policy as well as funder policies (important with doctoral centres with RCUK funding), and to teach students best research practice including data management planning.

As per Kevin's message, our approach has been to have a broad policy on research, covering all levels and all disciplines. However, while the policies are covered in a 'research students' handbook, we focus on doctoral students and have said we don't mandate DMPs for Masters students. Still, one supervisor has required it and others have said we should mandate it at MRes level too...

We don't mandate a timeframe other than "before data collection". The consensus from researchers was that there's too much variation for a set timeframe to be useful; the DMP is reviewed at all supervisory review meetings, with the form for the final meeting explicitly requesting confirmation of data storage at project end.

Interested to hear what others are doing, especially if extending to Masters projects or how much monitoring is in place."*