Research Data Management

When funders and journal publishers are asking you to provide a "public" link or "public access" to data underlying your research this means that they want you to deposit your data with a reputable repository. The National Science and Technology Council has put out a document on desirable characteristics of data repositories, as has the National Institutes of Health. 

An important note about "public" is that it does not mean available for anyone to download and use as they please. Public means that it is both Findable (via permanent URI such as a DOI) and the terms for making it Accessible and Reusable are clear such as terms of use, creative commons licensing, a formal proposal process and, potentially, legally binding data use agreements (The F, A and R in FAIR). It's important to note that repositories help manage access to data so that individual researchers do not need to respond to individual requests. 

Choosing what type of repository best suits your data ultimately depends on the research area, sensitivity of the data and the features of the repository. If it is also possible that if your research is funded, that your funder will indicate where the data should ultimately be deposited.

1. Human participants data may be sensitive depending on the dataset size, population, and topics covered (among other things). It's important to assess the risk of harm that inappropriate disclosure of the data could pose to the participants. With more than minimal risk, a repository with access controls should be considered. 
2. Data sensitivity may also be affected by other factors not related to human participants such as geolocations of endangered species or interests to national security. 
3. Disciplinary and data-type repositories often offer the most robust features for data in their collection purview (analytics, security, detailed metadata, etc.).

Please note that even controlled access repositories do require sensitive data to be de-identified as much as possible before deposit.