Share Data

Data sharing refers to making the research data underlying your publications available to others at low or no cost. Most federal agencies, such as the NSF and NIH, and many private funders now require researchers to share data publicly. 

We encourage researchers to use a data repository to share their data when possible. Repositories typically assign a DOI to your data and allow your data to be more easily found and cited. This helps your research find a wider audience and benefits other researchers in your field. 

What data should you share?

You should share the underlying data necessary to replicate the results of your research when possible. If sharing raw data, consider also sharing code or scripts used to process and analyze the data. Data should be accompanied by descriptive metadata; see Describe Your Data. 

Personally-identifiable information should never be shared. Data from human subjects research must be de-identified before sharing. Reach out to rdm@tufts.edu if you have any concerns about de-identifying your data. 

Choosing a Repository

There are two main types of data repositories: discipline-specific repositories, which accept data of a specific content or type, and generalist repositories, which accept data regardless of discipline. Generalist repositories should only be used when there is no applicable discipline-specific repository. Tufts provides institutional access to two generalist data repositories, Dataverse and Open Science Framework. 

Repositories are often free, but may charge fees to upload data. You can write these fees into your grant applications. 

Discipline-Specific Repositories

A small selection of discipline-specific repositories is below. Search for more in the Registry of Research Data Repositories or look in the Open Access Directory's list of data repositories. 

Journals or publishers may have data repositories they recommend, e.g, PLOS ONE, Springer Nature, ASM Journals. 

Science & Engineering

• NIH-supported data repositories includes 100+ disciplinary repositories, such as:
  • PubChem for data about chemical structures, annotations, and associated bioassay test results
  • GenBank for genetic sequencing data
  • OpenNeuro for MRI, PET, MEG, EEG, and iEEG data
• USGS acceptable data repositories
• IEEE DataPort accepts a wide variety of engineering data 
• Recommended repositories from PLOS ONE and ASM Journals

Social Sciences

• ICPSR for social and behavioral science data
  • Note: ICSPR allows for restricted-use data (for data that cannot be fully de-identified)
• Qualitative Data Repository for qualitative and multi-method research data

Tufts Dataverse

Tufts Dataverse provides a place for Tufts researchers to share their data. Researchers can self-deposit data at no cost, assign descriptive metadata, and receive a citable DOI for their dataset. 

Tufts OSF

Tufts is an institutional member with the Open Science Framework. OSF is an integrated research platform where you can work collaboratively, register projects, and share data. OSF is a good option for collaborating on a project with other researchers, and for sharing a wide variety of project materials (like presentations and posters). 

Generalist Repositories

These repositories accept data across all disciplines. See a comparison chart of the major generalist repositories.

• Harvard Dataverse (free, 1TB per researcher)
• Dryad (one-time fee of $150, 300GB per dataset)
• Figshare (free up to 20GB, up to 10TB+ with tiered fees) 
• Zenodo (free, 50GB per dataset)