STATUS: IN DEVELOPMENT
Short Description
“Data stewardship is the responsible planning and executing of all actions on digital data before, during and after a research project, with the aim of optimising the usability, reusability and reproducibility of the resulting data. It differs from data management, in the sense that data management concerns all actual, operational data-related activities in any phase of the data lifecycle, while data stewardship refers to the assignment of responsibilities in, and planning of, data management.” (Towards FAIR Data Steward as profession for the Lifesciences)
“Data stewardship is a relatively new profession and a catch-all term for numerous support functions, roles and activities. It implies professional and careful treatment of data throughout all stages of a research process. The core responsibilities and tasks vary, from policy advising and consultancy, to operational and technical support and IT related tasks. Responsibilities also vary between and among the different research-performing organisations, and data stewards (DS) often have different job titles.” (RDMkit)
Core of proper data stewardship are activities that ensure research data to be Findable, Accessible, Interoperable, and Reusable (FAIR) in the long term. This includes data management, archiving and reuse by third parties. Creating FAIR data requires attention from the planning phase of a scientific experiment to the life-long maintenance of the data. Hence, according to FAIRification models such as FAIRopoly and A Generic Workflow for the Data FAIRification Process, a FAIR data steward - with knowledge of the local environment and of the FAIRification process in general - should be part of your team, assuming this person will guide the FAIRification process. This could also mean that someone within the research team follows relevant training in order to take on specific data steward activities, and/or make use of existing (central) support data management services available in their own organisation or national organisations such as Health-RI.
Why is this step important
Good data stewardship has to adhere to the FAIR data principles and by definition implies long-term and sustainable care across multiple research cycles. For this reason, data stewardship is a collective endeavour, with actions (and competencies) needed from the individual researcher, from other scientists in the research project, from the institute, and also from the research discipline(s) and funders involved. Each organisation will implement its Research Data Management (RDM) policy tailored to its own situation and there will be a variation in how funders and institutional policies (or legislation) define roles and responsibilities, e.g. by placing differing expectations on the research team, host institution and third-party organisation (such as data centres) (Towards FAIR Data Steward as profession for the Lifesciences).
Having a FAIR data steward as part of your multidisciplinary team (see Metroline Step: Build the team) is beneficial for the following reasons (see FAIRopoly):
FAIR guidance: Having someone in the team with actual FAIR knowledge and expertise will help you reach your FAIR objectives (see FAIR Metroline: Define your FAIRification Objectives) more easily and faster.
FAIR standards: A FAIR data steward often has the required metadata standards and semantic modelling expertise, which is essential to achieving Interoperability between collections, studies and/or data(sets).
FAIR technologies: Additionally, a FAIR data steward often has extensive knowledge on technologies and tooling that help script or automate the FAIRness of data.
FAIR use-cases: As you don’t want to reinvent the wheel, it is the FAIR data steward that can link to the community’s good practices and uses case for FAIRifying data.
STAPJE EXTRA
1 Depending on the project, there may not be budget to hire datasteward for each research project. In my experience, only larger projects have the budget to hire a person for data steward activities only.
Within Amsterdam UMC, the central ‘data steward’ (Research Data Management) service can be consulted and specific requests that require more time may be invoiced. However, people from this department are not contracted to research projects. In addition, more often than not, data steward activities are embedded in roles of researchers. In this case, it is more valuable that researchers that also have data steward roles within their project, know where to find support within institutions and on a national level.
If specific knowledge is not present within our team, we look within our network and connect the researcher with people with the relevant expertise.
How to
Step 1
Consider what type of data steward your team would most benefit from. The NPOS/ELIXIR Data Stewardship Competency Framework distinguishes three data steward roles:
Policy oriented, focusing on policy development and the implementation of research data management practices in a team or organisation.
Research oriented, working directly with researchers and offering hands on support on data management issues.
Infrastructure oriented, translating the requirements of policies and science into suitable IT solutions, soft- and hardware, and tools
For these data steward roles, consequently, eight competence areas are distinguished (see below). Each task area requires different competencies. A single data steward can be responsible for all competence areas (including FAIR), but tasks can also be divided over multiple data stewards.
Make sure to consider the relevant competencies that can or should be covered by your team and to be aware of available resources for roles and/or competencies that fall outside the scope of your team.
Policy & Strategy
Compliance
Alignment with the FAIR data principles
Services
Infrastructure
Knowledge management
Network
Data archiving
Step 2
Discuss where in the organisation your data steward is allocated. Is it possible to (part-time) hire a data steward from a generic team (such as a central library, knowledge hub or IT team), or do you most benefit from a dedicated, project embedded data steward? Alternatively, you may want to consider consulting a dedicated data steward for missing competence areas in your team.
Depending on the project, there may not be budget to hire a dedicated data steward for a research project. In that case, you may consider using existing data steward support services in your organisation to guarantee that data stewardship is covered in your project.
Additionally, it is also useful consult data stewardship expertise in the broader network, on a local (e.g., the AUMC Data Stewards Network), regional (e.g., your regional Open Science Community), or even on a national (e.g., the Data Stewards Interest Group; Health-RI).
The following report on professionalising data stewardship in the Netherlands provides further information on both step 1 and 2 (chapter 2.3).
Step 3
Hiring your data steward following the formalised data steward profile adopted in the Netherlands, contributes to professionalising the role of the data steward in the Netherlands and strengthening the career perspective of your data steward. More information on these profiles in the formal Dutch job classification systems can be found in the earlier cited report (Annex 5 and further).
Policy & Strategy: design strategies for raising awareness of RDM policies and regulations
Compliance: advise on institutional compliance with RDM policies and regulations
Facilitating good RDM practices: advise relevant stakeholders on good practices of management of research data
RDM services: propose, implement and monitor RDM workflows and practices
Data infrastructure: identify the requirements for adequate RDM infrastructure and tools
Knowledge management: determine the adequate level of RDM knowledge and skills
Network & communication: create and participate in (inter)national RDM networks
Data sharing & publishing: analyse gaps in support for data sharing and publishing
Coordination of work: lead, supervise and support less experienced colleagues
Coaching & Process improvement: make proposals for improving work processes at different levels
Soft skills: this area comprises activities like accuracy and persuasiveness
Step 4
When extending your team with a FAIR data steward, also take the following into considerations:
Funders more and more often require a data steward to be part of a project team. Check for specific responsibilities and tasks in the grant proposal and make sure your team’s FAIR data steward is able to meet them.
Make sure data is handled in compliance with the journal and institute policy. Discuss in an early stage the potential condition of the journal you aim to publish in and the FAIR requirements by your institute. Extend your team with the required FAIR data stewardship knowledge and skills. These could relate to for instance uploading your (meta)data to a certain repository or catalog.
Expertise requirements for this step
To decide on the position of a FAIR data steward in your team, you need to have a clear overview of the added value of this role/function in relation to the FAIRification objectives of your project/department/team.
Practical Examples from the Community
European Joint Programme on Rare Diseases (EJP RD)
Has various FAIRification services, guidance, tooling and training, supported by a FAIRification stewards service of six people. The FAIRification stewards’ activities include, for example (see for more information the EJP RD FAIRification website):
Participate in meetings to exchange knowledge and experience and develop FAIRification guidance
Identify FAIRification bottle-necks and help apply the FAIRification process
Identify training needs and organize workshops and hackathons
Authors / Contributors
Sander de Ridder; Jolanda Strubel; Bruna dos Santos Vieira; Mijke Jetten; Fieke Schoots; Ines De Oliveira Coelho Henriques; Shuxin Zhang; Alberto Cámara; César Bernabé; Joeri van der Velde; Meriem Manaï