4 Challenges for institutional research data management support
Let's be honest, it's not always easy to find your way in the jungle of research data management services. Because we hope to learn from each other, Ghent University (Belgium) shares 4 challenges1 towards the development of it's institutional research data support and what they learned along the way2.
1.Where in the world to begin?
The challenge of starting something: Policy work
Chances are you don't need to start from scratch. Research data management is integrated in the research workflow in many different ways. Customs and guidelines might already be established for quality control, secure storage, interoperability and sustainable preservation.
But in order to move from a diffuse and unsystematic practice to a structural and university wide execution, you'll need a good basis.
A Research Data Management Policy can be a good foundation to start from as it will make the aspirations and goals set for the curation of data clear. It can also give an indication of where the university stands with the Research Data Services (RDS) and provide the necessary grip to start developing services for the curation of data.
At Ghent University a diverse working group (involving the central library, research coordination, ICT, the valorisation office, a representative of all faculties, the research director and the information security adviser) on Research Data Management was established to make sure the guidelines included the multiple challenges regarding good data management: from storage space to privacy over open access and data management planning practices in different scientific fields.
→ Our tip: Start to walk before you try to run. While open and FAIR data is an ideal we strive for, we started with guidelines and advice, supported by services rather than imposing requirements. The implementation of the policy is gradual and the autonomy of faculties and researchers to develop their own RDM initiatives is safeguarded. Although open research data is fostered, the policy recognizes that access may need to be carefully managed in order to maintain confidentiality, privacy, consent terms, security and to keep costs low.
2. Don't know much about history, don't know much biology…
The challenge of disciplinary research data and the need for skilled staff: Roles, Staff skills and expertise
The successful development and implementation of RDM depends on a wide set of skills. A university-wide survey of (doctoral) students and researchers on research skills, showed a lack of knowledge about the newest insights concerning data management and a shortcoming in experience with drafting Data Management Plans (DMPs).
To judge the quality of a DMP, peers and disciplinary staff with the expertise of the research field are best suited to validate choices made in the DMP. Yet faculty staff and fellow researchers are not always familiar with the overall reasoning behind DMPs let alone have the time to review them. To involve more people, deploy existing skills, combine efforts and learn from each other's practices, a small group of (aspiring) DMP reviewers was brought together to share their experiences on reviewing DMPs.
When the DMP-reviewing group started, it sought inspiration and good practices of other institutions and the use of an analytic DMP reviewing rubric came to their attention3. An analytic rubric is a standardized form to evaluate data management plans and can be used as a metric for compliance. The group developed two reviewing rubrics, openly available: one generic template and one based on the H2020 research data requirements. The rubric not only helped to standardize the reviewing process but also brought staff quicker up to speed with how to review a DMP without the need to be an expert in the research field.
Recently we recruited Research Data Management stewards who can further help with discipline specific queries.
→ Our tip: While discipline specific knowledge is irreplaceable, pulling resources and standardized procedures can help bring staff up to speed. Tools like a DMP reviewing rubric can standardize a procedure so the process becomes more efficient.
3. Where to look for guidance?
The challenge of providing the right research data services: Tools and services
Research Data Services (RDS) are not a single service and partnership is deemed essential to the success of the deployment of data initiatives.
Ghent University chose a "hybrid" approach: both involving faculties as central administration; in particular the central library, and looked at university-wide as well as faculty-level solutions.
First, the consultative services were prioritised. Providing information and sharing expertise is a great first step to achieve the take-up of good RDM for various reasons. It is relatively easy to do since the channels and expertise are readily available or easy to find and the implementation doesn't require a big budget.
The overview of guidelines, requirements and available services on the general website, was combined with practical, hands-on information on pages known as "researchers tips" which answers specific questions with practical tips and examples.
Additionally, to help researchers develop their data management plan, a localized version of the DCC tool DMPonline was developed. Dmponline.be was developed in cooperation with other Belgian universities and research organizations. A shared Belgian DMPonline platform has the advantage that common DMP templates (e.g. for external Belgian funders) can be added, new releases can be processed for multiple institutions at once, and new features can be co-developed while it still has every possibility to add local and institutional specific DMPs or guidelines.→ Our tip: Be pragmatic. What can we do with the tools, time and budget we have? We chose to prioritize consultative services since it would give us the opportunity to cover a basic need fast, effective and cost-efficient. Customized solutions as well as unified services provide a flexible framework that can adapt to the wide variety of challenges presented. And were it's opportune, we work together with other research institutions.
4. How to stop worrying and love RDM?
The challenge of engaging researchers and staff: Outreach and advocacy
Creating the right kind of environment for collaboration, sharing of knowledge and expertise rather than a one-sided top-down approach is a key part of the deployment of RDS for our institution. If data management is purely approached as a funder requirement, RDM in its own right as part of good research practices will be overlooked and researchers might thereby fail to adequately engage with RDM. It is therefore important to strike a balance between issuing guidelines and providing support to highlight the advantages of good RDM. We believe that training of RDS-related skills to cater to the needs of scientific researchers, can foster a collaborative environment that will enable the support for the RDM mission of the university.
To this end we have a compiled a programme of workshops, webinars and informative sessions around RDM, DMPs,services and funder requirements.
Furthermore national as well as international engagement for Open Science, provides the university with the advantage of obtaining information on the latest developments of RDM as well as access to the expertise of a broad network of Open Access experts and international connections with the Open Science community.
→ Our challenge: Advocating Research Data management and Open Science as more than an administrative task to be checked off.
→ Our tip: To spread awareness a repeated effort to be active, emphasizing the need and advantages of good RDM on different occasions, and in different ways and venues was favored rather than imposing a binding policy with obligations. National and international collaboration strengthens expertise and helps with visibility and consolidation of research data management as an integral part of responsible research practice.
All images by Patrick Hochstenbach CC-BY
1 To be honest, we had more than 4...
2 This blogpost is an updated and shortened adaptation of a previous article: https://citaliste.rs/casopis/br30/hermans_emilie.pdf
3 Initially the DART project developed a methodology for an evaluation rubric for DMPs. It was presented by the Digital Curation Centre in a practical exercise on reviewing DMPs at the international event: Supporting Data Management Plans: examples from national pilots and institutional projects (25 February 2016). http://www.dcc.ac.uk/events/workshops/supporting-and-reviewing-data-management-plans (accessed January 2017)
When you subscribe to the blog, we will send you an e-mail when there are new updates on the site so you wouldn't miss them.
