Best Practices: Choosing a Repository

The type of repository you use will affect the findability of your data. Use of a generalist repository can facilitate cross disciplinary discovery of your data. Using a discipline specific repository will make your data discoverable to other researchers in that field.

• Funder Supported, such as NIH
• Discipline-specific, such as the social science repository ICPSR
  • NIH database of domain-specific repositories
• Generalist, such as Dryad
  • NIH database of generalist repositories

Not all repositories are made equal. Review this list of desirable characteristics before choosing your repository: Desirable Characteristics for All Data Repositories

• Does your grant funder specify a repository? Done! That is the one you will use.
  • Not sure if your funder has a required repository? Check your funder’s requirements with SPARC
• If not, look for a discipline-specific repository:
  • Search re3data.org
  • NIH-supported domain specific repositories.
  • The Qualitative Data Repository
    • The Qualitative Data Repository is open qualitative data from all disciplines.
• Still no? Use a generalist repository. These are repositories that accept data from all disciplines.
  • Generalist Repository Comparison Chart
  • Dryad - UB is a Dryad member and therefore has some advantages:
    • Dryad provides data curation services.
    • For small datasets (under 10GB) your data publishing charges may be covered. Larger sets will incur fees.
  • Open Science Framework (OSF is also a great collaboration tool).
  • Other NIH-supported generalist repositories.

• NNLM Data Repository Finder
• Generalist Repository Comparison Chart

• Will your data require controlled access?
  • Controlled access means that anyone trying to access your data must obtain your permission first.
    • Some of the reasons to use controlled access:
      • Your data is sensitive. This is data that may contain personally identifiable data about human subjects or protected populations.
      • Your data must comply with laws and regulations such as HIPAA.
      • Your data contains intellectual property.
• Review the submission requirements prior to choosing a repository – make sure that your data is a good fit.
• Does the repository require a certain data standard? If so, make sure you structure your data with that standard.
  • For more information on data standards: Data Standards from the Office of Data Science and Sharing (ODSS)
  • If no data standard is required, consider using Common Data Elements.
• Does your chosen repository have fees for sharing or curation?
  • These fees can usually be written into your grant budget.
• Does the repository provide a DOI or other persistent identifier? This is important for discoverability and citation.
• If you haven’t already, register for an ORCiD. You will need an ORCiD to put your data in a repository.
• How long does the repository promise to save the data?
• What licensing does the repository require? For example, Dryad requires that you share your data with a CC0 license. See our page on best practices for copyright and data licensing.