On 17 March 2018 UCT released a new Research Data Management policy, with the stated purpose of “transform[ing] the way research is conducted at UCT by accelerating discovery, increasing the value of research decision-making, and catalysing changes throughout the economy and society that are of value to all citizens … [by] “ensur[ing] consistent research practice related to data management principles that support effective data sharing, including open access; and the need for data to be discoverable, accessible, reusable and interoperable to specific quality standards.”

In seeking to align UCT’s research data management activities with international best practices, this policy has profound implications for how researchers conceive of their data. In brief, one of the main goals of the policy is to shift research data management from a closed process managed internally according to standards set by the researcher or research group, to a more well-documented, public process in which data is conceived of as a potential output in its own right.

Any university-wide policy has to be able to apply across the institution, but in doing so necessarily has to trade disciplinary specificity for generalisability. Data management strategies and research data itself vary widely across disciplines, with some issues (such as High-Performance Computing, or qualitative de-identification) being unequally distributed across faculties or departments. As such there is always room for individual units and departments to create their own specific guiding documents that align with the spirit of the institutional policy but address the specific needs of their research context.

One such unit is the Gender Health and Justice Research Unit (GHJRU). The GHJRU, based in the Faculty of Health Sciences, undertakes research on topics related to gender and sexuality in the SADC region, with a particular focus gender-based violence, sexual and gender minority rights, and reproductive rights. In response to the policy, the unit decided to revise its existing data guideline document outlining a departmental data strategy that balanced the push for more open data practices presented in the institutional RDM policy with the needs of the GHJRU, in particular the need for tight oversight of data publishing given the need to protect the marginalised communities in which the unit works.

The GHJRU engaged with the Digital Library Services (DLS) unit in order to ensure their data management guidelines were in line with the new institutional policy. During June and July, the departmental representative and the data curation officer from DLS had a short series of face-to-face meetings and contact via email in which the departmental data curation plan was refined and new clauses entered to align more closely with the RDM policy, although little work needed to be done as the plan was already in an advanced state. Given the sophistication of the plan, as well as the specific curation needs of the unit (particularly in regard to the de-identification, curation and secure storage of the disclosive qualitative data common to GHJRU studies), the data curation officer created a specialised Data Management Profile template for the unit on DMPOnline. This template, tailored for the GHJRU’s needs, incorporates specific guidance drawn from international DMP templates, relevant UCT policies (the RDM policy, the Open Access policy and the Intellectual Property Policy) and the GHJRU guidelines. GHJRU thus became the fourth unit (alongside the Clinical Research Centre, the Libraries and Information Science Centre, and the Centre for Higher Education Development) to create its own DMP template.

Addressing the push for open, FAIR (Findable, Accessible, Interoperable and Reusable) data will be a challenge for UCT’s students and researchers. But the policy is not a monolith, and the GHJRU example shows how a research unit can actively engage with the policy on its own terms, drawing on its existing expertise to develop RDM strategies that work in its own context.

Departments or research units interested in creating their own DMP templates should contact the DLS Data Curation Officer at Thomas.king@uct.ac.za.