Why should I deposit my data on a repository?
An open access repository stores digital objects and makes them available (with their descriptive documentation and metadata) and downloadable. It’s accessible and interoperable through a OAI-PMH protocol and it deploys a long-term archiving policy.
Depositing and sharing your data on a digital repository is necessary to guarantee its findability and accessibility. Indeed, if you find a proper FAIR compliant research data repository you will also enhance the interoperability and reuse of your data (for other researchers, but also for your future-self!), thus ensuring the verifiability and reproducibility of your research results. Moreover, making your data available and accessible (even confidentially or upon request) via open repositories is one of the requirements increasingly demanded by funders and scientific journals which are more and more requesting a data statement attached to the publications. Importantly, also National Research Evaluation Agencies on the basis of the commitments defined by the Coalition for Advancing Research Assessment (CoARA) (which was signed both by UNIMI and ANVUR) are expressly requesting researchers to include datasets published in repositories among research products for evaluation purposes.
Repositories can be divided into: discipline-specific ones, institutional ones and generic ones. Check this selection flowchart or use the diagram on the left to navigate in the choice of the most suitable repository for your data:
- If your data contain sensitive information that cannot be anonymized, you should deposit it on a controlled access repository; for instance, it can be the case of heavy genomic data, which can be securely deposited on archives like the European EGA.
- If your data do not manage sensitive information or you are able to anonymize it, it is advisable to check if there is a specific repository for your discipline. Different tools can be used to conduct this investigation: OpenDOAR, FAIRsharing and re3data (the registry of research data repositories, whose use is also recommended by the European Commission).
- If none of the repositories you found meet your needs (or if you did not find any), use the research data repository of you institution, such as Dataverse UNIMI, the FAIR research data repository of the University of Milan.
- If your institution does not have a data repository, use a generic repository: check the Generalist repository comparison chart to navigate in the diverse possibilities.
Whilst browsing for the most suitable repository for your data, always bear in mind to check the following criteria and whether the repository has the following characteristics:
- Is the long-term preservation of data and metadata guarenteed? Look for the condition of long term preservation.
- Does the repository assign/include Persisten Identifiers (PIDs)? For example DOI or handle for datasets.
- What is the default license? For example in Dataverse UNIMI the default license is CC 0 but other licenses can be applied on datasets.
- Is the repository certified? For example, Dataverse UNIMI has obtained the CoreTrustSeal which guarantees the trustworthiness of the repository itself and the quality of th deposited data.
- Is there support staff to assist and help you when using the repository? For example, Dataverse UNIMI has a full-time support staff, check how to contact us here.
- Are there any storage costs involved? Costs should be taken into consideration amongst the allocation of resources in your Data Management Plan.
- Does the repository have a clear policy and easy findable terms and conditions? It is Important for granting access and reuse.
- Is there a descriptive and standard metadata set accompanying the datasets? Metadata make data findable, interoperable and understandable.
- Where is the location of the repository? It is important for the data protection law.
- Are there any size limitation on data files? These may cause you problems if you have heavy data files.
Finally, you can also consider to publish your dataset in a peer-reviewed data journal: this will enhance the transparency of the adopted methods and results
Example of generic data journals:
Examples of disciplinary data journals: