Publishing and sharing data under an open access model, which allows immediate, free, permanent, and unrestricted access to scientific outputs, is increasingly supported globally by both scientific institutions and funding providers. Making data accessible in accordance with the FAIR principles does not imply unrestricted access to data without any limitations. The goal is to follow the principle:
“...as open as possible, as closed as necessary.”
Research data, in line with this principle, should be published to the extent that maximally facilitates their reuse. On the other hand, data disclosure may be restricted, but only to the extent necessary to protect the rights and interests of the recipient, provider, or third parties. These protected rights and interests typically include the right to personal data protection, state security, or the interest in monetizing and utilizing research by the institution that conducted it.
According to Zákona 130/2002 o podpoře výzkumu, experimentálního vývoje a inovací (§2 Definitions, Paragraph 2, Letter o), research data are defined as:
"Information in electronic form that is collected or created during research, used as evidence, or generally accepted by the research community as necessary to validate findings and results."
The European Open Data Directive defines data as follows: "Research data includes statistics, results of experiments, measurements, observations resulting from fieldwork, survey results, interview recordings and images. It also includes meta-data, specifications and other digital objects. Research data is different from scientific articles reporting and commenting on findings resulting from their scientific research. For many years, the open availability and re-usability of scientific research data stemming from public funding has been subject to specific policy initiatives."
Research data should adhere to FAIR principles. These principles describe how data should be processed to ensure they are Findable, Accessible, Interoperable, and Reusable. Data should use standard formats, be accompanied by metadata, and include persistent identifiers (DOI, handle).
It is not contrary to the FAIR principles if access to data is conditional upon meeting certain requirements (e.g., signing a contract or adhering to contractual restrictions), provided these restrictions are necessary, transparently explained (e.g., in a DMP), and justified.
Although scientific articles themselves are not considered research data, they should also comply with FAIR principles, particularly in terms of findability and accessibility. It is strongly recommended to implement FAIR principles from the beginning of the research process.
A detailed description of the FAIR principles can be found in English on the Go FAIR website which is also linked in the following overview of the FAIR principles.
For data to be reusable, it must be possible for both humans and machines to find them. Machine-readable metadata are key for this purpose. Each output should have a unique and persistent identifier, such as a DOI (Digital Object Identifier), which ensures that data can be reliably located even after a long time. Artistic research outputs, such as images, audio tracks, or multimedia projects, must be accompanied by high-quality metadata to facilitate their discovery. Persistent identifiers like DOI ensure long-term findability not only in academic databases but also in cultural and creative industries.
Data and outputs must be accessible to users, even if access restrictions apply (e.g., due to data sensitivity). This is usually managed through clearly defined licenses and appropriate access control. It is crucial that metadata remain accessible at all times, even if the content itself is subject to restrictions. In artistic research, this could mean ensuring access to multimedia outputs through licenses such as Creative Commons, which clearly define usage rights for users.
- A1. (Meta)data are retrievable by their identifier using a standardised communication protocol
- A1.1 The protocol is open, free and universally implementable
- A1.2 The protocol allows for an authentication and authorisation procedure where necessary
- A2. Metadata should be accessible even when the data is no longer available
To ensure data can be used across different systems, they must be recorded in compatible and standardized formats. The use of open data formats (e.g., XML, JSON) and clear metadata standards facilitates easy information exchange. In the field of artistic research, this may include using standards such as Dublin Core for describing artworks, scripts, or musical compositions, enabling data sharing and processing across different platforms.
Outputs should be as reusable as possible, which means clear documentation, consistent formats, and licenses that allow for repeated use. Providing context and information about data provenance is also essential. In artistic research, this includes sharing datasets of images, audio recordings, or videos with corresponding metadata that provide sufficient context, allowing other researchers to build upon previous work.
Here you can find a simple self-assessment tool for evaluating FAIR data. It allows you to assess the "FAIRness" of your data and determine how you can better implement the FAIR principles.
As open science practices gradually become more integrated into academic work, the requirements for handling research data are also increasing. A key tool designed to facilitate the management of research data is the Data Management Plan (DMP). While DMPs are most commonly used in the natural and social sciences, they are also essential for the humanities and artistic research. Grant providers (such as the Czech Ministry of Education – OP JAK, GAČR, or TAČR) require them as a mandatory document at various stages of a grant project.
A DMP serves both as a reporting document for the grant provider (making it an additional administrative requirement for researchers) and as a living document that helps researchers anticipate and address challenges related to research data management. These challenges may include copyright issues, GDPR compliance, data backup, and systematic data labeling. A well-prepared DMP can increase the impact and reach of research outputs.
What Counts as Research Data? In natural and social sciences, defining research data is relatively straightforward—it includes measurement results, survey responses, databases, tables, graphs, and more. In the humanities and artistic research, however, researchers may not always work with such structured datasets. What, then, should be considered research data in these fields? Research data encompasses any inputs used in the research process. This may include primary and secondary literature, scans of archival documents, photographs, databases, spreadsheets, lists, research notes, or audio recordings. Since grant providers require that all relevant data be published alongside research, researchers must determine what should be made publicly available. This requires balancing relevance, copyright considerations, and data sensitivity. Generally, data should be published to an extent that allows the research to be verified. Some data (such as archival documents) may not be publishable, while others (such as literature references) may not need to be. The DMP serves as a planning document where researchers clarify their data management strategy and communicate it to the grant provider.
If you are considering applying for grants such as OP JAK, GAČR, or TAČR, it is advisable to appoint a data steward within your team. This person will be responsible for drafting and updating the DMP and managing research data. Salary costs for data stewards are typically eligible expenses in grants, so they should be considered during grant writing. For further assistance, please contact us at open.science@amu.cz.
Compiled based on KNAV (CC-BY)
Academy of performing arts in Prague
Malostranské náměstí 259/12
118 00 Praha 1
Registered number: 61384984
VAT number: CZ61384984
Data box ID: ikwj9fx
