Data Steward

The role of a data steward is fairly new in research and thus is still evolving and being defined. In 2019, the final report for the Dutch project titled Towards a community-endorsed data steward profession description for life science research defined data stewardship as follows:

“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.”

From this definition, it follows that a data steward is concerned with, in broad terms, ensuring the quality of the research data throughout the research data life cylce by applying the FAIR Data Principles.

Tasks and Skills

A data steward's task can take on various forms, often depending on the exact position within the organization and the individual's personal skill set and background. The duties may involve defining appropriate repositories for data publishing, consulting on data formats and metadata standards, setting up data documentation and metadata management systems, e.g. through an electronic laboratory notebook (ELN), and training researchers on best practices. The Engaging Researchers with Data Management: The Cookbook elaborates on the various tasks and skills for the role of data steward. In it, Alastair Dunning, the Head of Research Data Services at TU Delft, is quoted saying:

“Interpersonal skills are key to the success of the Data Steward.”

This reflects the importance of a data steward's communication skills.

Data stewards may work directly at an institute, but, in Germany, many are part of larger collaborative projects such as Collaborative Research Centers or even part of various NFDI consortia. They often take on the role of communicating a specific research community's needs and requirements, enabling developers at IT centers, university libraries, ELN providers, and the NFDI to create tailored solutions for digital research data. In the above-mentioned cookbook, several strategies for developing and settling data stewards are outlined. One of the presented models is the “rent-a-data-steward”, which differs from the current models. Here, data stewards are not permanently located within institutes, instead, they are “loaned” to the institutes, projects, etc. for a certain period of time.

To efficiently manage data and maintain an overview, data stewards can use a Data Management Plan as a supporting tool.

Recently, a BMBF-funded project titled DataStew has been analyzing the concept of data stewardship in the German academic landscape and identifying various stakeholder requirements to be able to provide recommendations for education and training for the role. Similar to the above-mentioned Dutch project, the first results here show the multi-faceted task areas assigned to data stewards. From consultating on general RDM services, development or expansion of IT-tools and infrastructure for RDM in research, discipline-specific consulting, conducting their own data-focused research for specific disciplines, to occupying coordinative roles, one data steward position is rarely like the next.

Data Stewards in Chemistry

Chemical research is often a combination of observational data and digital data. While the focus in RDM and data stewardship typically lies on the latter—as emphasized in the definition quoted at the beginning of this article—the oberservational aspect cannot be ignored. Traditionally, this information is recorded in hand-written lab journals. Here, data stewards must understand the type of laboratory work carried out and assist in selecting appropriate ELNs to digitize this type of information. For digital data, data stewards must be aware of community-specific standards for data formats and minimum information (MI) regarding metadata . With this knowledge, they can assist in efficiently implementing data workflows that automate data transfer from device to secure storage locations, data format conversions, and metadata annotation. For example, Chemotion ELN's device integration can facilitate such workflows, while directly connecting the data to molecules and reactions recorded in the ELN. Such processes allow for a well-documented dataset to be published in repositories alongside the results reported in journal articles.

Networking and Continuous Training

Due to their wide breadth of duties, data stewards are continuously involved in networking, exchange, and furthering their own knowledge on RDM approaches, best practices, tools, and services. The RWTH Aachen University, for example, holds a monthly Open RDM Network Meeting (in German) geared at data stewards, while state initiatives such as fdm.nrw offer various networking and training opportunities. NFDI4Chem also organizes an entry-level workshop series titled FAIR Research Data Management: Basics for Chemists for researchers and data stewards alike and focuses on research data management concepts and tools in chemistry.