DMP Best Practices

These are best practices to consider while working on your DMP. They are based on the most common stumbling blocks we see DMP authors encounter. If you want to see examples of real DMPs, the DMPTool has put together this list of high-quality DMPs. (But remember Best Practice #1 below: your funder’s requirements are an important part of this process.)

1.  Look up the DMP requirements for your grant proposal.
Funders typically outline or link to their DMP requirements in their funding announcements. If you are writing your DMP for a grant proposal, it’s important to begin by reviewing your funder’s requirements and structuring your DMP accordingly

2.  Anticipate software and storage needs by considering the types of data that will be created.
Some instruments need unique or proprietary software and some projects generate large amounts of data (either in number of files or file sizes). Anticipating these issues will make sure there’s no scrambling.

3.  Create or adopt standard terminology and file-naming practices.
Decide on conventions and stick with them. Document your standards so transitions are smooth as graduate students, post-docs, and collaborators cycle on and off projects.

4. Set a schedule for your data management activities.
For example, nightly back-ups onto shared storage ensures more than one copy of the data exists. Monthly file and/or directory clean-ups will help keep “publication quality” data safe and accessible. (Note that ideal backup and cleanup cycles will vary depending on the project.)

5. Assign responsibilities.
For example, assign a data manager who can check that backup clients are functional, monitor shared directories for clean-up or archiving maintenance, and follow up with team members as needed.

6. Think long-term.
For example, data associated with publications needs to be preserved long-term. Decide where it’s going (e.g. a discipline or institutional repository) and when it’s going there. Other types of data may warrant this level of preservation as well (e.g. reference, irreplaceable, long-term monitoring, etc.).