FAIR Data and Open Data
The preservation and accessibility of research data helps to increase the reusability of once collected data and to ensure reproducibility of research results. Research data should therefore be findable, accessible, interoperable and reusable (FAIR Data).
In particular, meaningful metadata, persistent identifiers and unique licenses are required to make data reusable for humans and machines and thus FAIR. FAIR Data are not necessarily Open Data: they can also have limited access, but it is most important to describe them as precisely as possible.
Unrestricted availability of research data (Open Data) aims to prevent duplication of existing data and to allow subsequent use of existing data. Many publishers and journals demand research data underlying publications to be published, for example as supplementary material, in order to ensure the reproducibility of research results. A growing number of funders also ask for Open Access to research data generated in funded projects.
Generally, research data are published by depositing them in a suitable repository. In addition to subject-specific repositories, there are interdisciplinary repositories such as Zenodo; a central directory of repositories is re3data.org. AUSSDA (for social sciences) and GAMS (for digital humanities) are subject-specific repositories at the University of Graz that currently has no institutional research data repository in place.
Funders such as the FWF and the European Union (within the framework of Horizon 2020) expect that research data generated within funded projects will be made freely accessible. Costs and expenses for data management and data publication can be claimed within the proposal. If research data cannot or should not be made freely accessible, the reasons must be specified. The reasons named most frequently for "opting out" are IPR protection, privacy, and a jeopardisation of the project objectives by Open Data.
If research data is to be made freely accessible, the legal preconditions (copyright, licensing, data protection) must first be examined. The applicable data protection regulations must be observed whenever handling any personal data. In order to ensure legal clarity for subsequent use, research data should be provided with a free license (e.g. Creative Commons).
If research data are reused, guidelines such as the Data Citation Principles must be observed in the interest of good scientific practice.