What Is Data Sharing?
Data sharing generally refers to the process of making the underlying data necessary to replicate the results of research to the public or to a research community.
Just as importantly, data sharing is not the same thing as:
- Publishing an article. An article typically contains limited expressions of the data collected, such as charts, but these expressions are not sufficient to replicate the article's findings.
- Retaining your data. While the University of Kentucky Data Retention and Ownership Policy requires the retention of research data for five years, retention refers to maintaining your data for review or auditing purposes. Separate steps must be taken to share your data.
Why Share Your Data?
In addition to the fact that many funders now expect data sharing to take place, there are a number of benefits to sharing the data associated with your research:
1. Increase citations of published research. High-quality datasets with robust documentation and their own DOIs can be independent research outputs (separate from an article) and offer another opportunity for citation in the work of other researchers.
2. Increase research efficiency by avoiding duplication of effort.
3. Promote scholarly rigor by making methods, protocols, and data more readily available for peer review and scrutiny.
4. Enhance visibility and scope for engagement, including 'citizen science' and public engagement beyond the academy.
5. Enable new research through the re-use of data and support the aggregation and re-analysis of data from a wide range of sources.
6. Enhance collaboration and community-building across disciplinary, institutional, and national boundaries
7. Increase the economic and social impact of research, innovation in business and public services, and the return on public investment in research.
How Can Data Be Shared?
Below are the principal ways in which researchers share their data along with a list of benefits and drawbacks for each.
| Method | Benefit(s) | Drawback(s) |
|---|---|---|
| Project Website |
|
|
| Journal or Data Paper |
|
|
| Available Upon Request |
|
|
| UKnowledge |
|
|
| Discipline-Specific Repository |
|
|
Preparing Your Data for Sharing
Selecting a Repository for Your Data
While there are other options for sharing data, repositories are a popular option because they do not require ongoing maintenance from the researcher and they make datasets more easily discoverable. You can deposit your data in UK's institutional repository, UKnowledge, or choose an external repository.
Desirable Repository Characteristics
The NIH offers guidance on the selection of a data repository. Some of the most important attributes that a repository should have are:
- Unique persistent identifiers (such as a DOI). These are crucial to ensure that your data is accessible and discoverable.
- Long-term sustainability. The repository should be able to maintain your data for an extended period of time and have a plan to protect deposited data in the event of a system failure.
- Curation and quality assurance. The repository should have policies in place to ensure that data comes from legitimate research and that the associated files can be opened and are complete. (Note that curation and quality assurance are not the same thing as peer review. Data curators evaluate the technical quality of datasets, not the scientific validity of any associated findings or their reproducibility.)
- Free and easy access. Datasets should not be locked behind paywalls and should be easy for repository users to locate.
- Retention policy. The repository should have clear language documenting the length of time that they guarantee to maintain deposited data.
UKnowledge
UKnowledge captures, stores, organizes, and provides open and stable worldwide access to UK's intellectual capital. It facilitates reuse of deposited materials to the extent warranted by copyright law or by the licensing terms of the concerned material. Members of UK's academic community are encouraged to contribute their scholarship to UKnowledge.
UK Libraries is a member of DataCite and registers Digital Object Identifiers (DOIs) for datasets deposited in UKnowledge. DataCite Research enables keyword search for datasets (with accompanying information) hosted in repositories worldwide. This mechanism helps other researchers find your research data.
Domain-Specific Repositories
Sharing your data in a domain-specific repository makes it more discoverable for researchers in the same field. Additionally, it is more likely that a domain-specific repository will have enhanced support, such as a tailored metadata schema, for your data. Grant funders may require you to deposit in a specific repository and will assist you in identifying one. The following lists can also be used to find repositories:
- NIH-supported domain-specific repositories
- Registry of Research Data Repositories (re3data.org)
- List of domain-specific data repositories (including in the humanities) by Tri-College Libraries
Generalist Repositories
If there is not a discipline-specific repository appropriate for your data, consider depositing it in a generalist repository. Generalist repositories allow researchers from any discipline to host and share their data.
If you are new to generalist repositories, begin by exploring those that participate in the NIH's Generalist Repository Ecosystem Initiative:
You can easily compare offerings using the Generalist Data Repository Comparison Chart.
Some repositories, such as Dryad, charge data processing fees for the cost of data storage and upload. These fees are sometimes covered by journals publishing articles associated with the data. You may also include costs associated with data sharing in your budget on many grant applications.