The Rebecca J. Holz Series in Research Data Management is a monthly lecture series hosted during the spring and fall semesters. On April 24, Clare Michaud and Morgan Witte, two M.A. students at the UW-Madison iSchool, spoke about key topics to address in a data management plan (DMP) for NSF funding and how to use DMPTool, a free, online tool for creating funder-compliant DMPs.
To start the talk, Michaud went over best practices for writing an effective DMP for NSF funding. Major funding agencies are increasingly requiring grant applicants to submit a DMP along with a project proposal. A DMP is a brief document (no more than two pages!) that outlines how data will be managed over the course of a research project and once the project is over. This has been a requirement for all project proposals to the NSF since 2011, and each of their seven research directorates have slightly different expectations about how to address the topics that should be covered in DMPs.
NSF’s research directorates include Biological Sciences; Computer and Information Science and Engineering; Education and Human Resources; Engineering; Geosciences; Mathematical and Physical Sciences; and Social, Behavioral, and Economic Sciences. In general, applicants to any of these research directorates should address:
- data types and how data will be stored and backed up,
- data documentation, and the formats and standards for their data and metadata,
- data sharing and access,
- data archiving and preservation, and
- data management roles and responsibilities.
Though all NSF grant funding program proposals require information about data management, not all researchers seeking funding will necessarily be creating or managing data during their research project. This is especially common for applications to the Division of Mathematical Sciences (within the directorate Mathematical and Physical Sciences). When this is the case, researchers just need to submit a statement that no data management plan is necessary. However, researchers whose projects may not include data management should still consider addressing how they will manage the outputs from grant funding, like articles or conference presentations and include information about that in the data management plan of their proposal.
Addressing roles and responsibilities when managing data within a lab or research group is something that is increasingly required across funding agencies, and this is an area that NSF requires all applicants to address. Some key aspects of data management that benefit from role-and-responsibility assignments are with data storage and backup, data organization, and data archiving and preservation.
During the second portion of the talk, Witte demonstrated how to use DMPTool to create a DMP according to what is required for researchers submitting proposals for funding from the Social, Behavioral, and Economic Sciences directorate. DMPTool offers templates for DMPs that comply with requirements for most major funding agencies, making it an easy-to-use resource to guide researchers through the writing process. RDS recently enabled DMPTool’s Feedback function, making it easier for the UW-Madison.
DMPTool also allows researchers to make their DMPs openly accessible to other DMPTool users at their institution. This is an excellent way to facilitate data- and research-sharing practices, and to give colleagues examples of DMPs (though we need to note that these are not vetted by UW-Madison, RDS, or DMPTool). It’s important to note that, while DMPTool allows researchers to draft and export well-formatted DMPs, it does not submit DMPs to funding agencies.
All of the materials from the talk are available on GitHub, including the slides, a handout on using DMPTool, and links to resources for what each NSF directorate expects in a DMP.
For additional information on creating a DMP, check out our updated guidance.