Table of contents
1. Scope
The goal of this policy is to offer clear guidelines for research data management (RDM) in line with relevant legal and policy frameworks (appendix B).
RDM refers to the responsible organisation, storage, archiving, and publishing of research data during all stages of the research process.
Research data are all data, digital and non-digital, that were generated or reused during the research process and used to form scientific conclusions. This encompasses information that is used to form and test hypotheses upon which conclusions are based. This definition of research data includes observational, experimental, simulated, derived, and compiled data. It may include software generated during the research process or the workflow describing the research process. Where applicable, research institutes should specify what research software falls under the definition of research data.
This policy applies to all research data collected or reused by employees and all persons affiliated with the Radboud University.
Each research institute at Radboud University has its own RDM policy that follows the general guidelines outlined below. The director of research [1] is responsible for the development, establishment, and revision of the research institute’s RDM policy. The research institute’s policy is intended to be a practical implementation and further elaboration of the general RDM policy.
2. Why RDM?
Careful organisation, storage, archiving, and publication of research data is important for many reasons, such as:
- Verifiability of research results, necessary for research integrity
- Reproducibility and replicability of research, which are important for research quality
- Increased potential for reuse, ensuring efficient and effective use of research resources
- Improved prospects for recognition of good research practice and accountability
- Appropriate management of privacy-sensitive information, contributing to ethically responsible and legally compliant processing of personal data in research
- Compliance with policies, including (inter)national and regional regulations, funding bodies, and scientific journals
Radboud University sees it as its responsibility to make research data as publicly accessible as possible. In line with the open science principles, the motto is as open as possible, as closed as necessary – a view also held by many funders in the Netherlands and across Europe.
2.1 What is FAIR?
The FAIR principles [2] are guidelines to improve findable, accessible, interoperable, and reusable. A more detailed explanation of the FAIR principles can be found in appendix C.
Although the FAIR principles encompass four dimensions, this policy focuses on the findable and accessible aspects. The way to achieve this is outlined in this policy.
3. RDM in research
3.1 RDM for researchers
The following sections cover the guidelines for RDM for which the researcher is responsible.
Data management plans
A Data Management Plan (DMP) is mandatory for all research projects that generate new research data, or that reuse existing data in ways that require additional research data management decisions, risk assessments, or agreements.
Radboud University requires researchers to create a Data Management Plan (DMP) for all research projects that generate new research data or reuse existing data in ways that require additional research data management decisions, risk assessments, or agreements. A DMP ensures that data are well-organised, secure, and aligned with FAIR principles, thereby strengthening the quality and impact of the research. It also supports researchers in estimating RDM-related costs and incorporating these into funding applications. In addition, a DMP facilitates early consideration of ethical approvals, privacy assessments, and any required legal or contractual arrangements. A DMP must be prepared before the start of data collection or data acquisition and must be kept up to date throughout the project.
Data storage & sharing during research
Research data must be stored in an adequate facility where at least one other member of the research institute can get access
While research is being carried out, research data must be stored at a facility that is adequate in terms of availability (the data may not inadvertently be lost), integrity (the data may not inadvertently be modified), and confidentiality (the data may not inadvertently be made available to unauthorised persons). Data must be stored in such a way that, apart from the researcher, they can be accessed by at least one other authorised member of the research institute. This is to prevent data loss when the researcher leaves or is unavailable.
Radboud University provides storage facilities that meet these requirements. The research institute’s RDM policy specifies which of these systems are recommended at their research institute.
Data archiving & publishing
Selected research data must be archived and published for the purpose of verification, replication, and reuse. Radboud University requires researchers to follow the three guidelines on data archiving and publishing outlined below. These guidelines must be met no later than the publication date of the first scientific publication based on the research data. In the case of data reuse, the researcher should not archive or publish research data that have been properly archived elsewhere, unless new information is added or multiple datasets are combined in an innovative way.
Archiving and publishing data must in all cases comply with contractual agreements with funders and legal and ethical guidelines, such as intellectual property rights, user terms from data suppliers, and privacy legislation. The researcher is responsible for ensuring that agreements with collaborators and informed consent forms allow for meeting the three principles.
1. Research data underlying a scientific publication must be archived in a trusted data repository for at least 10 years after the publication date
Purpose: To allow for the validation of research findings presented in scientific publications.
Elaboration: Research data that are required for the verification and replication of published research findings must be preserved for a minimum of 10 years after the publication date of the related scientific publication. After 10 years, it should be evaluated whether it is useful to continue to preserve the data. Data should be removed only after careful consideration. The research institutes decide by whom and how this evaluation is done.
To facilitate this evaluation, Radboud University encourages researchers to document the reason for archiving (e.g. “to allow for verification of published research”, “valuable dataset for reuse”, or “unique dataset”).
Important considerations for the selection of other research data that are to be preserved for the long-term, or to continue to preserve data after 10 years, are:
- Usefulness of the data for future internal or external reuse
- Uniqueness of the data
- Cost of data collection
- Protection of the privacy of research participants
- Costs and energy expenditure of data archiving, and thus the size of the dataset
Archiving must be done in a trusted digital data repository. Technology may change considerably in 10 years’ time. This might render data inaccessible if they are not curated over time. Trusted data repositories have a mission to provide continued access to digital data, now and in the future. A further definition of and explanation about trusted data repositories can be found in appendix D. Recommended trusted data repositories are specified in the research institute’s RDM policy. In rare cases, large, privacy-sensitive data may be stored for the minimum retention period in an internal institutional storage location. The research institute’s RDM policy specifies whether and to which data such an exception applies.
The above applies to all digital research data, including digitised versions of non-digital research data. The retention period for non-digital research data and any requirements for the digitisation of non-digital research data are specified in the research institute’s RDM policy.
Research data acquired during a research project that do not underlie a scientific publication, may also be preserved for the long term.
2. Research data underlying a scientific publication must be archived as open as possible, as closed as necessary
Purpose: To allow for efficient reuse of research data and for the verification and replication of published research.
Elaboration: Research data that are relevant for the verification and replication of published research must be made publicly available, unless there are valid reasons not to do so.
Other research data, which are not relevant for the verification and replication of published research, should only be made publicly available if they are expected to have value to the scientific community or society at large.
Valid reasons not to make research data public include:
- Ethical treatment and protection of the privacy of research participants (when deidentification is not (sufficiently) possible)
- Preventing harm in the case of sensitive information about companies or socially sensitive information (e.g. about specific communities)
- Knowledge security
- Public safety
- Intellectual property rights of third parties
- Contractual obligations with research funders
In these instances, access to data may be closed, restricted, or temporarily closed (i.e. placed under an embargo).
If research data cannot be made public, the related metadata must be made publicly available (under a CC0 licence) to ensure findability of the data. Most trusted data repositories automatically publish metadata in the “public domain” (i.e. under a CC0 licence).
It is advised that any scientific publications linked to closed, restricted, or embargoed include a statement explaining what data exists, why access is restricted or closed, and -for restricted data- by whom and how access can be obtained.
3. Research data underlying a scientific publication must be archived according to the findable and accessible principles (F and A of FAIR)
Purpose: To allow for efficient internal and external reuse of research data and for the verification and replication of published research.
Elaboration: Research data that are relevant for the verification and replication of published research must comply with the findable and accessible principles. The interoperable and reusable principles may be relevant in specific contexts and disciplines, but are not part of the current policy framework. Other research data, which are not relevant for the verification and replication of published research, should meet the findable and accessible principles if they are expected to have value for internal reuse or to the scientific community or society at large.
To meet these criteria, research (meta)data must be archived in an appropriate trusted data repository (see appendix D) that assigns a globally unique and persistent identifier (PID - such as a DOI) to the data. The research institute’s RDM policy provides guidance on selecting a data repository. Minimal criteria for data repositories are included in appendix D.
Research data must be archived with sufficient metadata and documentation to ensure that data can be found, interpretated, and used. Metadata must include a link to the related scientific publication, and should include a link to other related works (datasets, software, publications, pre-registrations, etc.) if applicable.
Research data that are published must be made available with clear access conditions and terms of use. Without this, a dataset cannot be reused since a data user does not know if and how data can be used. To allow for efficient reuse of scientific data, Radboud University encourages researchers to release research data under a CC0 or CC-BY licence (or equivalent) whenever possible.
A published dataset’s metadata must remain public even if data are retracted or removed after the expiration period. This is often handled by the repository. This is important since even if the original data are missing, it can still be useful to track down the associated people or publications. Furthermore, it is important to confirm that the dataset once existed and what happened to it.
Where applicable, the research institute’s RDM policy specifies which domain-relevant open science standards apply.
Radboud University strongly encourages researchers to link research data publications to their ORCID [3], so that all of a researcher’s research output can be unambiguously linked to them. Some funders also require the use of PIDs for researchers, such as ORCID.
Registration of data publications
Registering published research data, preferably linked to the researcher’s ORCID, is strongly encouraged
Radboud University recognises the importance of research data publication. By registering data publications, researchers enable accurate reporting, ensuring that data is acknowledged and valued as a form of research output. This process not only enhances transparency and reproducibility, but also contributes to the recognition of data as a vital component of academic contributions.
Radboud University therefore aims to have all published research data registered in the CRIS (current research information system). The research institute’s RDM policy specifies how the registration of published datasets takes place. The registrations should use the PID and link to the associated publication(s). Further, use of the researcher’s ORCID is strongly encouraged.
Data reuse
Radboud University researchers are encouraged to reuse research data whenever possible
Radboud University encourages that its researchers reuse research data whenever possible. Reusing data can save time and money, reduces the burden on participants, and fosters the open science principles.
When reusing data, Radboud University researchers must cite said data in any scientific literature based on the data. This is to ensure that researchers are acknowledged for producing research data. Citations should allow for identification of, access to, and verification of the specific data that supports a claim. Citations must at least include title, author(s), publication year, persistent identifier, publisher, and if applicable: version number and resource type (e.g. dataset, software).
When reusing data, the researcher must respect any intellectual property rights and comply with the user terms.
3.2 RDM for PhD candidates
In addition to the requirements for RDM for researchers listed above, PhD candidates must follow the RDM requirements specified in Radboud University’s Doctorate Regulations.
3.3 Policy implementation
Data stewards
Each research institute has one or more data stewards. Data stewards are the first line of support within the research institute regarding RDM. Another core task of the data steward is to actively contribute to the implementation and revision of the research institute’s RDM policy. An extensive description of the data stewards’ tasks and responsibilities can be found in Radboud University’s Tasks & Responsibilities in RDM.
Support and training
Radboud University provides knowledge, support, and training to assist researchers. The Digital Competence Centre (DCC, division Information & Library Services) serves as a central hub and knowledge centre for researchers as well as data stewards. A detailed overview of the DCC’s roles can be found in Radboud University’s tasks & responsibilities in RDM.
Infrastructure
Radboud University provides (access to) sufficient infrastructure that enables researchers to comply with this policy and requirements from funders, scientific journals, and open science principles.
This includes:
- A dedicated tool for data management planning
- Systems and support for the storage, backup, and safe sharing of research data during the active stage of research
- A registration tool for research data
- A data repository and support for the long-term preservation and FAIR publication of research data
4. RDM in education
Research data pertaining to bachelor’s or master’s theses are archived at a facility that is adequate in terms of availability (the data may not inadvertently be lost), integrity (the data may not inadvertently be modified), and confidentiality (the data may not inadvertently be made available to unauthorised persons). Radboud University provides (access to) a system that is suitable for this purpose. Education institutes are free to use this system at their own discretion.
The retention period for research data pertaining to bachelor’s and master’s theses is equivalent to the legal retention period for theses, i.e. seven years.
Within the bachelor’s and master’s programmes, attention is paid to RDM in a manner that is appropriate to the discipline and the phase of the study programme. Bachelor’s and master’s theses include a statement on the management of the research data pertaining to the thesis.
It is also possible that research data pertaining to bachelor’s or master’s theses are part of the research that has been carried out by a researcher or project manager. In this case, the policy principles for RDM in research apply and the supervisor (i.e. the researcher) is responsible for adherence to this policy.
5. Monitoring and compliance
RDM policy compliance is monitored at the level of the research institutes. Each director of research is responsible for the implementation of a research institute’s RDM policy which complies with this policy. Furthermore, they are responsible for monitoring compliance. The research institute’s data steward is responsible for having a basic overview of compliance with the RDM policy.
Radboud University’s DCC monitors and evaluates the provided RDM support services and infrastructure to ensure that they remain useful.
The director of education is responsible for implementing and monitoring compliance with this policy insofar as it relates to academic higher education.
6. Appendices
Appendix A: Glossary
Archiving: archiving research data is about preserving research data for the long-term, by placing the data in an archive/data repository. Archiving may include the publication of the archived data.
Data management plan (DMP): a dynamic document developed at the start of a research project which outlines how data will be collected, processed, analysed, described, preserved, and shared.
FAIR data: acronym for Findable, Accessible, Interoperable, and Reusable describing data principles. An elaboration can be found in appendix C.
Licence: this is a legal arrangement between the creator of the data and the end-user, specifying how data can and cannot be reused.
Metadata: descriptive information that facilitates cataloguing of data and data discovery. Metadata allow humans and machines to more easily understand and interpret data.
Open science principles: promote open and collaborative research which leads to greater scientific and societal impact.
Persistent identifier (PID): a unique and stable denomination (reference) of a digital resource (e.g. research data) through allocation of a code that is persistently and explicitly referenced on the internet by a service, e.g., a DOI or handle.
Publishing: publicly disclosing research data.
Research software: source code files, algorithms, scripts, computational workflows, and executables that were created during the research process or for a research purpose.
Trusted digital data repository: A service that provides long-term care for and (public) access to research data. An elaboration can be found in appendix D.
Appendix B: Legislation and regulations
This policy framework is premised on existing frameworks, which include (but are not limited to):
- The general data protection regulation (GDPR)
- Uitvoeringswet algemene verordening gegevensbescherming (UAVG)
- The dutch copyright act, database act, and patent act
- Medical-ethical research protocol requirements pursuant to the dutch medical research (human subjects) act (WMO), as evaluated by the medical research ethics committee (METC)
- Netherlands code of conduct for research integrity
- Radboud University’s doctorate regulations
- Radboud University’s knowledge security guidelines
- Radboud University’s guidelines on control of research data
- Radboud University’s tasks and responsibilities in RDM
Appendix C: FAIR research data
Note: Although the FAIR principles encompass four dimensions explained below, this policy focuses on the findable and accessible aspects.
FAIR is an acronym for findable, accessible, interoperable, and reusable. The FAIR data principles [4] state that it should be possible to find research data, there should be information about how to gain access to them, they should be compatible with other data, and it should be possible to reuse data. The principles are not necessarily about open data: data can be archived in a FAIR manner without being available to others.
Findable
(Meta)data should be discoverable by humans and machines. To a large extent, this is achieved by placing (meta)data in a trusted digital repository. The data are referenced with a persistent identifier (e.g. DOI or handle) and the metadata include the identifier of the data.
Accessible
It should be clear how humans and machines can gain access to data. Access should be possible using open and free protocols such as the internet. Metadata should be accessible, even when the data are not.
Interoperable
(Meta)data can be exchanged and used, now and in the future, across different platforms, for example by using open file formats. Data can be integrated with other data using metadata standards, ontologies, and controlled vocabularies.
Reusable
Data should be reusable by others. Therefore, the data should be documented and licensed. Information on the content, context, and structure of the dataset should be included. A machine-readable licence should be included, so that re-users know what they can do with the data.
Appendix D: Trusted digital data repositories
A data repository is “an archival service providing long-term care for digital objects with research value” [5]. A trusted digital data repository has been defined as having “a mission to provide reliable, long-term access to managed digital resources to its designated community, now and into the future”.
In practice, this means that a trusted digital data repository ensures that research data are not merely safely stored, but remain accessible and understandable to end users over a long period of time. The repository accomplishes this by monitoring and acting on changes in technology and its user community.
Trusted data repositories meet a set of requirements. At a minimum, Radboud University requires that Radboud University researchers use a repository that:
- Assigns a persistent and globally unique identifier (such as a DOI) to each dataset
- Has a continuity plan to ensure ongoing access to and preservation of the research data
- Guarantees the integrity and authenticity of the data
- Ensures that datasets are indexable and searchable
- Provides data access to authorised users via an open, free, and universally implementable protocol (for example over the internet)
- Provides (open) access to data and releases data with clear user terms/ a licence
- Ensures that metadata remain public, even when data are removed or retracted
Describes data with rich, standardised, machine-actionable metadata, at least consisting of:
a. The persistent identifier of the dataset
b. Title, author(s), publisher, publication year, version number, resource type, subject, and a description of the content and context of the research
c. Link to the related publication
d. Data usage terms (licence)
[1] At Radboudumc, the responsibility for RDM policy rests with the heads of departments. Throughout this document, references to the director of research can therefore be read as referring to the head of department in the context of Radboudumc.
[2] Wilkinson, M., Dumontier, M., Aalbersberg, I. et al. The FAIR Guiding Principles for scientific data management and stewardship. Sci Data 3, 160018 (2016). https://doi.org/10.1038/sdata.2016.18
[4] Wilkinson, M., Dumontier, M., Aalbersberg, I. et al. The FAIR Guiding Principles for scientific data management and stewardship. Sci Data 3, 160018 (2016). https://doi.org/10.1038/sdata.2016.18
[5] Andreu, T., Anglada, L., Antos, D., Bähr, T., Brzeźniak, M., Burgi, P.-Y., Cavet, C., Celjak, D., Crépé-Renaudin, S., De Loof, C., Dillo, I., Dubois, O., Fernandes, R., Forsström, P.-L., Ganis, G., Gibney, E., Holl, A., L'Hours, H., Lamers, D., … Wyns, R. (2023). EOSC Preservation: Overview Discussion Paper. Zenodo. https://doi.org/10.5281/zenodo.7516259