Diversity of data stewardship within a federated research infrastructure

Diversity of data stewardship within a federated research infrastructure

As the requirements for research data and its handling are constantly changing, the range of tasks and topics in scientific institutions is also becoming more diverse and larger. This poster will outline the diverse developments, challenges and difficulties surrounding the topic of research data in a federated, decentralized research infrastructure such as the Helmholtz Association from a cross-institutional and cross-disciplinary perspective. As interest in managing research data increases in many parts of the research ecosystem, policies and infrastructures are being implemented in more and more contexts.

Data management teams can consist of data stewards, librarians, software developers, and scientists from various research areas. They create conditions to ensure digital reproducibility and best practices for research data. Data stewards collaborate with researchers in a research ecosystem on different levels. In a large scientific organization, they can be deployed for specific research disciplines, priority initiatives, major projects or in infrastructure-related services. Data stewardship includes responsibility for data and coordinating the necessary activities for data exchange processes, including the collection, storage, backup and use of data[i] also with regard to licensing agreements and ethical standards. Therefore, data stewards assist researchers in practical applications, enabling scientists to benefit from research processes.

It is a challenge to make different data sets from the scientific process of various disciplines findable and reusable. A central tool is the use of metadata to enrich research data with standardized and machine-readable information. To coordinate this process, central infrastructures such as the Helmholtz Metadata Collaboration (HMC) and the Helmholtz Federated IT-Services (HIFIS), bundle scientific expertise from various research areas on the subject of metadata or build and maintain an IT infrastructure for networking research areas and centers. For instance, they offer consulting services, provide information about suitable technical solutions, and ensure uniform scientific standards. These structures create synergies and increase the efficiency in the use of resources and knowledge throughout the research association. With these supporting infrastructures, data stewardship can benefit from a variety of services and developments, related to culture change in research data provision and reuse. At the same time, it is an overarching goal to provide research data with metadata and make it FAIR (Findable, Accessible, Interoperable, Reusable)[ii]. Here, the data stewards can support the digital workflow from data generation to documentation to analysis and archiving by using appropriate interfaces or linking via persistent identifiers (PIDs) on an infrastructural level.

An additional challenge in the field of data stewardship is the wide range of developments at national and international level with regard to a standardized approach to data. Data stewards also act as an interface between scientists and scientific support units, such as infrastructures internal to the institution (IT department or libraries) or even to national infrastructures such as the National Research Data Infrastructure (NFDI).

Guidelines and policies were drawn up to coordinate the handling of research data and software in the Helmholtz Association. The requirements for the research data cycle depend on the discipline and also differ in the individual research areas. With the ongoing digitization of research and teaching, the number of research software solutions for scientific knowledge acquisition is constantly increasing. In many cases, the provision of appropriate software is of great importance for the reproducibility of data analyzes and for the subsequent use of the respective research data. For this reason, suitable referencing and provision of the software used or developed is essential. The requirements for a data steward for research software also include licensing issues. In this case, cross-research consulting offers can support the work of a data steward.

In addition, facets of open science will be considered as an essential component of FAIR data stewardship. In order to organize open science according to the principles of openness, transparency, quality assurance, networking and sustainability, researchers will be supported in the implementation by providing infrastructures, services, consulting and training. In perspective, central open science infrastructures such as publication platforms, in-house publishers, repositories and consulting services are to be established and expanded. The requirements for a cultural change with regard to open science are set out in the Helmholtz Open Science policy[iii]. A data steward plays a central role in communicating with scientists in particular and can accompany cultural changes.

In closing, current developments on research evaluation and incentives play a central role in the transformation of the scientific enterprise. For an established, sustainable research data management (RDM), the recognition of the research performance of scientists is another important component. In this context, the evaluation of new indicators for data and software products is necessary. Hereby, pure metrics are not a solution, quality issues should be central. In general, data stewardship can support research assessment in several ways. Data stewards can provide insights into the researcher’s data management practices and help them identify areas where additional support is needed[iv]. At the same time, they can help establish best practices for data management and develop services to support them. Data stewards can assist to evaluate the quality of data management plans (DMPs) and provide feedback to researchers on how to improve them[v]. In addition, they can support identifying areas where policies and guidelines need to be developed or updated. In their work, data stewards gain insights into the use of data infrastructures and can determine the need for services or tools that researchers require. They can work in both discipline-specific and general infrastructure areas, and embedding them in a data management team is one way to encourage transfer and foster changes in assessment practices.

The requirements for data stewardship should be considered in their diversity based on the federated, decentralized research association. Science supportive units offer a broad portfolio to support sustainable and FAIR RDM as well as export and publishing routines. The tasks of data stewards depend on the embedding in the scientific ecosystem. Data stewardship is intended to support cultural change in research data and software, as well as Open Science at discipline-specific or general infrastructure level.