Data Management Plans (DMPs)
Data Management Plans (DMPs) are required in DFG funding proposals since 2022 as well as for EU Funding Programmes 2021-2027. DMPs act as a reporting tool for funders to hold grant recipients accountable to conduct good and open science with periodic updates or upon changes. For researchers and other stakeholders, DMPs are meant as a living document that accompanies them from proposal writing or project start to the sharing of their data and findings.
Content of DMPs
In a DMP, researchers usually describe the data, their generation and processing during the project, as well as how the data and research results will be archived afterwards to remain available, usable and comprehensible. DMPs usually contain the following information:
- Responsibilities and obligations
- Description of the research project
- Costs and resources
- Description of the research data: type, quality, organization and usage
- Metadata and referenceability
- Publication (access and reuse)
- Data citation
- Storage and security
- Digital preservation
- Legal aspects and anonymisation
When implemented correctly, a DMP can benefit all stakeholders of a research project despite the initial overhead of creating the DMP itself:
- Transparency and reproducibility: besides serving as a reporting tool for funders and governing institutions, a DMP can be handed to new collaborators of a project to convey a short description, experimental design, methodological spectrum, proposed hypotheses and general progress. This way, all stakeholders can quickly look up the state of data, find the responsible person for questions and document their own contribution.
- Resource management: including a timeline of personal and lab resource availability can enable service facilities (e.g. sequencing or IT) to more accurately schedule their resources/capacity and project costs.
- Standardize processes: DMPs can easily be adapted to similar projects by researchers in the same field/institute to reduce the effort for new proposals and resort to fixed entities (e.g. ethical board).
Even though, it is generally possible to formulate a DMP in an office document, the use of more dynamic and machine-readable formats finally enables the full anticipated potential. In Germany the Research Data Management Organiser (RDMO) has gained widespread adoption among institutes and consortia. This web-tool is used to create institute-wide templates and organize DMPs in different versions and share them with all stakeholders.
RDMO organizes individual plans around predefined templates that reflect the requirements of the respective institution, discipline or funder. This ensures machine-actionable compatibility for administrative stakeholders and re-usability for researchers in following projects. These templates usually contain much more aspects that do not have to be answered right from the start of a project, but can be completed as the research progresses.
- Example of a good DMP: Molin, E. (2018). Behave Working Data-Management-Plan. Zenodo. https://doi.org/10.5281/ZENODO.1243717
- Biological & Environmental Sciences
- Health Sciences
- University of Minnesota (incl. School of Public Health): web page
- Clinical trials
- National Institutes of Health (NIH): download
- PAPA-ARTiS: download
- European Clinical Research Infrastructure Network (ECRIN): pdf (p. 48)
Machine-actionable DMPs (maDMPs)
- Michener, W. K. (2015). Ten Simple Rules for Creating a Good Data Management Plan. In P. E. Bourne (Ed.), PLOS Computational Biology (Vol. 11, Issue 10, p. e1004525). Public Library of Science (PLoS). https://doi.org/10.1371/journal.pcbi.1004525
- Miksa, T., Simms, S., Mietchen, D., & Jones, S. (2019). Ten principles for machine-actionable data management plans. In F. Ouellette (Ed.), PLOS Computational Biology (Vol. 15, Issue 3, p. e1006750). Public Library of Science (PLoS). https://doi.org/10.1371/journal.pcbi.1006750