requirements on data management planning and sharing.
Data Management 101: Planning Checklist
The National Institutes of Health (NIH) have had data sharing requirements for several years, but just this January
What type of data will be produced?
the National Science Foundation (NSF) began requiring
›› How will data be collected?
a data management plan as part of all new NSF proposals.
›› Will it be reproducible? What would happen
As Johnston has met with faculty, they have expressed interest in making sure their NSF applications have robust data management plans, believing that will give them an edge in a very competitive grant process. But not all faculty understand the value of creating a plan, or what makes a plan robust. The workshops that Johnston and Lafferty offer answer both the “why” and “how” of data management planning.
if it got lost or became unusable later? ›› How much data will it be? and at what growth
rate? How often will it change? ›› Are there tools or software needed to create/
process/visualize the data? ›› Storage and backup strategy? What standards will be used for documentation
and metadata? ›› How to document data collection?
Saving Time, Increasing Impact
›› Is there good project and data documentation?
For those who are not under a mandate to create a plan,
›› What directory and file naming convention will be used?
it can seem like a lot of extra work to do so. But Johnston
›› What project and data identifiers will be assigned?
explains how that effort early on can save time later. For
›› Is there a community standard for data
example, complete documentation for a data set provides evidence for the published results of research and also makes it easy to field requests from funders or other researchers seeking information about the data. Further, studies have shown that researchers who post their data to a public space like a website or repository see an increase in citations to their work. Having a plan to share data after publication of a researcher’s results can do more than stimulate citations of that
sharing/integration? What steps will be taken to protect privacy, security,
confidentiality, intellectual property or other rights? ›› Who controls it (e.g., PI, student, lab, University)? ›› Any special privacy or security requirements
(e.g., personal data, high-security data)? ›› Any embargo periods to uphold? If you allow others to reuse your data, how
publication. In fact, many data sets have value beyond
will the data be accessed and shared?
their original research. Take the Human Genome project for
›› Any sharing requirements (e.g., funder data
example: in 1990, an international research team set out to
sharing policy)?
sequence the thousands of genes that make up human DNA.
›› Audience? Who will use it now? Who will use it later?
By sharing their data throughout the project, they not only
›› When will I publish it and where?
finished the project two years ahead of schedule, but open
›› Tools/software needed to work with data?
access to this data continues to generate new research aimed at curing genetic diseases.
How will the data be archived for preservation
and long-term access? Being convinced of the value of planning for ongoing management of their data is only the first step. To guide researchers through the steps of creating an actual plan, Johnston and her colleagues have created a checklist (see sidebar). These detailed questions make clear that researchers need to consider how they are planning to use the data today, as well as how they or others might use
›› How long should it be retained (e.g., 3–5 years,
10–20 years, permanently)? ›› What file formats? Are they long-lived? ›› Are there data archives that my data is appropriate
for (subject-based or institutional)? ›› Who will maintain my data for the long-term?
it tomorrow. And because every researcher and every data set is different, Johnston and her colleagues stand ready to help them answer those questions.
adapted from the National Science Foundation’s guidelines by the University of Minnesota Libraries
*
LIB.UMN.EDU
7