Kim now has a good sense of how much of Dr. Smart's data can be shared without restrictions, and how much of it will need extended protections. She's presently at the point of investigating another aspect of data sharing and access - the means of distributing the data for reuse and creation of derivatives. Dr. Smart would like to be able to point to his data sets from this research, making them freely available, rather than have to provide them on demand. He's interested in making his data easily discoverable and easily accessible.
In this part, you will find out some of the ways to make research data available for sharing and reuse, such as modes and levels of access.
Earlier you learned about restrictions on data, especially if the data being collected are sensitive or confidential in any way, as well as about intellectual property rights that can be associated with data.
In a DMP researchers also need to state how they will share and make accessible data that are not restricted. Public access to data marks a central motivation behind the DMP requirement. As stated previously, data that have been generated by a federally funded project is publicly funded data - that is, data made possible by taxpayer dollars. Data sharing is also integral to the responsible conduct of research; it contributes to the verifiability of research results.Thus, in this part of a DMP, you describe how you will make your data available for sharing and reuse.
Using his own lab as an example, in the following video Andrew Stephenson explains some of the reasons why granting agencies, such as NSF, have decided to consider the data from funded research to be public data.
In thinking about how your data will be disseminated and shared, consider the primary community (or communities) of interest for your research - that is, who is the anticipated audience for your data?
Another factor to address is how the data in your field are currently shared. Are there well-known data repositories through which data of the type you'll be generating are typically made accessible? Before the DMP requirement was implemented, it was acceptable for researchers to make assurances in grant applications that their data would be stored on a hard drive and be made data available upon request. Since the mandate, it is recommended that you make your data publicly available via a repository service. (The next part of this tutorial will focus more on data repositories.)
Finally, the DMP should state what the intended uses of the data are likely to be, and who is likely to use the data. How might the data be reused and repurposed and thus transformed? What alternative uses of the collected data might be possible? As described by Dr. Traverse in the following video, it may not even be possible for you to imagine some of the future uses for your data!
The DMP should describe how the project will share data, making them publicly available (unless the data are restricted) for broad access and sharing. Providing access to data is responsible science and helps ensure verifiability of the research results.
It is worthwhile thinking how data in your field are typically shared and accessed. Perhaps there are certain data repositories that hold data sets in your field that you and your lab team frequently consult. Increasingly, the practice for researchers is not to keep their data on a hard drive and distribute it on request but to deposit them to a repository, linking to them from publications and laboratory websites, so that they can be discoverable and findable.
The DMP also typically states when the data will be made available, whether during the project or after it. If embargoes will be implemented, then this fact needs to be surfaced as well, with accompanying information about any time constraints. Similarly, will all the data be shared, or only some of them? If the DMP hasn't addressed levels of access (e.g., who is permitted access to the data), then this is where that information should be given, too.
True or False: The current best practice for data sharing is to state in your DMP that your research data is available upon request.
(a) True
(b) False
ANSWER: (b) False. Increasingly, funding agencies wish to see in DMPs that researchers will share their data via data repositories, or an institution-wide repository service like Penn State’s ScholarSphere. By depositing data into a repository service, the onus is removed from the researcher in making sure the data are preserved over the long term and thus accessible on an ongoing basis. This way, researchers can post the permanent link or URL or DOI for the data set to the project website and be assured the data set will continue to be accessible.