Recommendation 1 - Encourage data citation practices
RPOs should encourage authors to cite data specifically in the same manner as detailed in the Specific Recommendations for Authors.
Guidelines and best practices disseminated by the RPO should recommend citing data specifically, providing at least the six core components of a data citation. Examples of this could be requirements or recommendations in open science policies, web content guidelines, course materials and instructions for students, PhD handbooks or formal requirements for theses and report series, etc.
Recommendation 2 - Encourage authors to report data publications when reporting their research outputs
RPOs should encourage data authors to submit information on data publications when actively reporting their research outputs to institutional repositories, bibliographic metadata catalogues, library catalogues, publication lists on faculty profile pages, and similar functions or infrastructures.
Authors should be encouraged to provide the six core components of a data citation as a minimum amount of metadata when reporting a data publication.
Recommendation 3 - Include data publication metadata in institutional repositories, publication lists and similar indexing services
RPOs should treat data publications as a valid research output and enable indexing of such publications in institutional repositories, bibliographic metadata catalogues, library catalogues and similar services. They should include at least the six core components of a data citation in the metadata for data publications. If possible, data publications should be distinguishable in entries and metadata by being treated as a unique resource type, such as “Dataset”.
RPOs should, if possible, make efforts to find and index past data publications using available metadata sources, especially if data publications have not been included in active research output reporting in the past.
RPOs should make sure that indexed data publications are reported together with other institutional research outputs to national research publication indexes, national CRIS or grant tracking systems, national library catalogues, and similar infrastructures if relevant. The reported metadata for a data publication should include at least the six core components of a data citation.
If possible, information about additional contributors and organisations, related publications, research project identifiers, funder identifiers and grant IDs should also be included in the reported data publication metadata, using machine-actionable PIDs when feasible. This will enable a large-scale overview and assessment of research outputs.
Recommendation 4 - Include and leverage the added value of metadata describing data publications in DMP tools, research information systems and similar administrative infrastructures
Research project management and administration conducted by RPOs may make use of project indexing support systems, such as current research information systems (CRIS), also known as Research Information Management Systems (RIMS). RPOs may also use other tools for creating reusable project-related structured metadata and documentation, such as tools for creating data management plans. When possible, RPOs should facilitate the production and inclusion of data publication metadata, encourage citing data and aid data publication flows using such support systems.
Examples of this could be using existing project metadata to automatically create pre-populated metadata entries for data publications to be completed at a later date, requesting details about FAIR data publication planning at an early stage, and allowing metadata input of at least the six core components of a data citation when prompting for information about planned or published new data publications. Secondary data used in the project should likewise be included in the metadata using at least the six core components.
By including data publication metadata in tools and support systems at an early stage, project members will be encouraged to relate to and actively and systematically cite existing data as well as new data throughout the project. This may also streamline the data publishing process and aid in decision-making. At the same time, requirements for making data publications FAIR will be highlighted and early planning will be promoted.
Support systems may also aid data publication workflows by providing access to a selection of project metadata in advance to be used by external infrastructures, such as data repositories in preparation of publication entries, or metadata aggregators enabling scientific knowledge graphs.
Recommendation 5 - Take data publications into account when performing institutional bibliometric analysis or creating output summaries
RPOs often use bibliometric analysis or other output summaries as a part of the process of mapping research outputs of institutions, departments or research groups. Whenever possible, they should consider data publications as a valid research output and take them into account when analysing or summarising outputs.
Data publications should be included in this process to provide a more complete basis for research assessment and reporting. Taking data publications into account is also likely to encourage data producers and authors to publish formal data publications as well as to cite data when relevant. RPOs are key stakeholders for creating and maintaining a sustainable data citation culture.