Publishing data sets enables data re-use, validation, and attribution. Data sharing is one of the topics covered in Data Management Plans.
Published research data can be used to validate the research results, can contribute to new research, and can earn attribution and citations.
High profile data sets that are able to support a large body of research can be published in specific data journals like Scientific Data. Peer review will be done on the data level in this case.
Most data sets do not undergo peer review before publications. They are published to underpin research results presented in peer-reviewed journals, as an increasing number of journals require the availability of data.
Zenodo is a multi-disciplinary repository funded by the EU that can be used to publish all sorts of research results. Zenodo accepts data sets as well as publications, presentations, posters, multimedia, software, or educational resources.
Upload on Zenodo is not restricted and any object deposited will get a DOI as a persistent and unique identifier (Learn about identifiers). With the DOI, you and others can cite the data set in a publication. Some journals require the availability of the underlying data before accepting articles for publication. Uploading data on Zenodo would meet this requirement.
The University of Limerick has a research community on Zenodo. If you upload material into Zenodo, please do link to this community (this is an option during upload): https://zenodo.org/communities/ulir/.
The Registry of Research Data Repositories is a global registry that can be used to find data repositories. It covers repositories from different academic disciplines and can be browsed by country or by content type. It presents repositories for the permanent storage and access of data sets to researchers, funding bodies, publishers, and scholarly institutions.
The ideal file formats for data publications are:
If a particular software or specific code is needed to open or understand data this should be explained in the documentation. Links to software and code (for example published on Github) should be provided if they are not widely known. Data can also be published in multiple formats to have for example a version with specific features in a proprietary format alongside an open format for preservation and ease of use in the future.
Recommendations on suitable formats are provided by the Library of Congress.