In the United States, research data that is considered factual cannot be copyrighted. However, sometimes the associated metadata, databases, figures, software, or work that could be considered a “creative” output can be considered an asset you want to control reuse or redistribution over by applying appropriate licensing. Licenses define how others may interact with, reuse, modify, or redistribute your work. Choosing a license for your data ensures that it is used appropriately by other researchers.

For scientific or factual data, many researchers choose to apply a Creative Commons 0 license, so that it is clear the data is being distributed freely. At most a CC-BY (Creative Commons Attribution) license should be applied.

For creative works, there are varying levels of Creative Commons licenses that you can choose to apply. The license levels build on one another and range from unrestricted to fairly restrictive terms. The Creative Commons website has a good guide that can help you decide what restrictions you may want to apply including request for attribution or non-commercial use. To learn more about Creative Commons licenses and how they differ from copyright view Lesson 2 of the Copyright and Fair Use micro-course.

For software or code, there are multiple choices. You will want to select from a license you are comfortable from ones such as the GNU licenses, MIT license, or Apache licenses. For databases and their content, the Open Data Commons Licenses can be used for the licensing of databases and their contents. Databases are a more complicated situation in that the database may have copyright protection, but its data does not or may have separate copyright, we recommend reviewing Cornell’s guidance on this licensing.

You should cite datasets for the same reasons you cite books and journal articles: for dataset creators to receive appropriate credit for their work, and to make clear the antecedents to your research.

Data citation standards may vary between disciplines and some professional organizations, academic journals, and repositories may also have guidance on preferred data citation formats. However, in general, the information you capture in a data citation is similar to the information included in a citation for any other work.

The Inter-university Consortium for Political and Social Research (ICPSR) suggests the minimum elements of data citation as:

• Title
• Author
• Date
• Version
• Persistent identifier (such as the Digital Object Identifier, Uniform Resource Name (URN), or Handle System)
  • Note: Persistent identifiers are preferred, but if not available a URL will suffice.

As mentioned above, a persistent identifier is a useful piece of information for data citation. A commonly used persistent identifier for research data is a DOI, or a digital object identifier. A DOI is a series of alphanumeric characters that serve as a unique identifier for a specific publication, dataset, or other digital object. The DOI for that object won’t change over time as the URL might if the web page is moved or the website is changed. Instead, the URL is information that is attached behind the DOI and it can be updated over time.

This allows researchers to make their datasets easier to locate, access, and cite for other researchers. Reliable location, identification, and citation by others is a critical component for researchers to enable reuse of their data, replication of their work, and to track the impact of their research.

A DOI has to be provided by a DOI Registration Agency, however many publishers, repositories, and institutions work with such agencies in order to provide DOIs for their communities. When sharing your data, we recommend checking with your publisher or repository to see if a DOI is provided for you.

Storing and using sensitive data requires extra protections to ensure that not only are we complying with important legal obligations but also protecting the safety of human subjects’ information appropriately.

Here at UW-Madison, if you have sensitive or restricted data, you should only be using approved tools for that type of data. You should follow campus guidance regarding handling sensitive data and reach out to your departmental IT or DoIT for guidance on what approved tools are available to you for your data type.

The Research Data Services website hosts a comprehensive comparison of UW-Madison storage and backup options. A note here that some of these options are not suitable for storing sensitive data. Your departmental IT or DoIT can help you come up with appropriate storage and backup solutions for your data.

If you are looking at cloud applications for your data storage, backup, or computing, always read the terms of service. It is important that you understand the permissions you are granting to the company that supports the application and how any data might be potentially shared. Part of protecting your data is understanding the risks to your data – and that includes knowing what risks could come through your storage and backup tools.

For those going to school or working at UW-Madison, we recommend that you always use your institutionally provided accounts over your personal accounts (for example, Google Drive). UW-Madison has an agreement with these providers to provide more intellectual property protections than your personal accounts would provide.

A good rule of thumb to remember is LOCKSS, or Lot Of Copies Keep Stuff Safe. However, you don’t need to go overboard with the number of copies you have. Typically, the rule to follow is the rule of three: three copies, in at least two physically separate locations, on more than one type of storage hardware.This might look like:

• A copy in active storage. This is a copy you are regularly accessing and working on during your research. It will likely be on your computer or a lab’s shared network drive.
• A second copy on a different device on- or off-site, such as an external hard drive in your office or a backup server provided by your IT department
  • Note: Be cautious about using USB flash drives to backup your data. They have some advantages that can make them an appealing option: they’re affordable, they’re convenient, and you probably own at least one already. However, flash drives’ portability makes them easy to misplace, have stolen, or accidentally break.
• A third copy, preferably off-site. This might be on a cloud application like Box, Google Drive, or other appropriate cloud solution

The goal here is to make your backups and storage as physically far apart as possible to prevent any loss due to natural disaster, such as a fire or flood occurring in the lab where you’re doing research. If your backups are all housed together, it could ruin both the primary copy of your data and any backups that you keep in the same building. Having at least one off-site backup increases the chances that you can restore your data if a disaster happens.

One of the most important things for file naming is to develop a naming convention, a template of standard information you’ll use in most file names, and to always use that convention anytime you have multiple related files in a folder. Without a set convention, you may end up recording haphazard information, or not capturing enough important information, each time you create a file: this will make it harder to remember what keywords you can use to search for the file.There is not one recommended naming convention that will work for everyone. Each project and person are different, but below we’ve laid out some suggestions for creating a file naming convention.

For more information on how to create a file naming convention, visit Lesson 2 of the Research Data Management Micro-course.

Another key component of an effective file management strategy is establishing a well organized hierarchical folder structure to go along with your file naming conventions. While there’s no easy answer as to how many folders you should have or how to best organize them, the trick is to create a structure that balances breadth versus depth.

Try to limit the number of top level folders you create and try to limit the number of folders nested within those as well. You don’t want to create so many layers in your hierarchy that accessing the actual data files becomes difficult but you also don’t want to have too many files within each folder which may also make finding your data more difficult. There is a useful micro-course on research data management that covers in more detail hierarchical organization and folder best practices.