Data Management Plan (DMP) Guide

Biological Sciences

Highlights

• BIO is committed to timely data sharing and the DMP should address plans for how this will be achieved.
• The DMP should cover both data (digital and analog) AND physical materials (samples and collections).
• DMPs for Collaborative proposals need to cover all of the various data from each collaborator. 
• BIO provides a suggested organization for the DMP which is summarized below:
  1. Data types and formats (including specimens and samples)
  2. Standards for file formats, data, and metadata
  3. Roles and responsibilities
  4. Dissemination methods
  5. Policies for data sharing, public access, and re-use 
  6. Data preservation (archiving) and access throughout the lifecycle.

Computer and Information Science and Engineering

Highlights

• Clarification that "research data" includes original data as well as metadata (e.g., experimental protocols, software code written for statistical or experimental analyses or for proofs-of-concept, etc.).
• Provide details on the data retention and sharing periods and any limitations on data sharing.
• Mechanisms and formats for storing data and making them accessible to others, which may include third party facilities and repositories.
• Data documentation and metadata needed to use the data.

The guidance document provides a section titled "Additional Guidance on Selecting or Evaluating a Repository" (at the very end) which can help you decide where to deposit your data.

Education & Human Resources

Highlights:

• Two new definitions took effect in Jan 2018:
  • “Access to data” refers to data made generally available without  explicit request by a third party (e.g., data or material made available on a public website).
  • “Data sharing” refers to the release of data or material in response to a specific request from a third party.
• Roles and responsibilities related to DMP activities need to covered.

Engineering

Highlights

• Both physical and digital data may be covered but digital data is of more concern.
• Differentiate between restricted and released data.
• MINIMUM data retention period: 3 years post-award or 3 years post-public release, whichever is later.
• Basic compliance means releasing and preserving analyzed data and associated metadata.
  • Typically this data is published in research papers, theses/dissertations and/or as supplementary information (SI).
  • This should be treated as "bare minimum" compliance.

Note: The engineering guidance document contains a few sentences addressing research publications in the Data Formats and Dissemination section that is confusing. The intent is likely the following: if you have a good reasons to delay the publication of research results then this reason may also apply to the research data.

Geosciences

Geoscience has three sets of guidance documents, one for each division. Each set of guidelines is very different so make sure you know which one your proposal falls under.

Atmospheric and Geospace Sciences

This division has a PI advice document, a DMP form, and instructions for using the form.

• If data is to be distributed by a website or database managed by the research group then a statement of how long each will be in operation/maintained should be included in the DMP.
• If data is in physical form (such as lab notebooks) address physical safeguards against fire and water

Earth Sciences

• Earth Science values long-term, high quality data sets.
• Preservation is required for all data, samples, physical collections, etc. needed for long-term research and education.
• Data sharing should have low access barriers and only incremental costs (if any) and the data must be accompanied by data documentation.

Ocean Sciences (OCE)

This division has the longest guidance document. Read it carefully.Two things to note are:

• Data must be shared no later than 2 years after collection.
• Data should be deposited to the appropriate Data Center (see appendix).

Mathematical and Physical Sciences

The Mathematical and Physical Sciences directorate has five sets of guidance documents, one for each division. Each set of guidelines is different so make sure you know which one your proposal falls under.

AST and CHE differentiate "access to data" and "data sharing" and ask for PIs to cover both if applicable.

Astronomical Sciences (AST)

• The guidance document provides items and examples to help the creation of astronomy DMPs.
• Address both "access to data" and "data sharing" if applicable.

Chemistry (CHE)

• *Updated June 2017* - The new document provides much clearer expectations and examples. Changes include:
  • DMPs should not be generic. Specifics should be provided and the DMP should be written with the needs of the project, and it's data, in mind.
  • "Simply putting data in Supplementary Material of a publication is not sufficient data management."
• Address both "access to data" and "data sharing" if applicable.

Materials Research (DMR)

• DMR provides no extra guidance but specifies that the DMP should address aspects of data retention/sharing related to questions about published results and that members of formal collaborations can refer to existing policies and practices.

Mathematical Sciences (DMS)

• For many proposals, as long as a clear justification is provided a statement that no DMP is needed will suffice.
• Those that require DMPs will likely be interdisciplinary projects. In these cases the DMP should address the data practices of the other disciplines.

Physics (PHY)

• PHY provides no extra guidance but specifies that the DMP should address aspects of data retention/data sharing and that members of formal collaborations can refer to existing policies and practices.
• If there is no data related to published results this should be stated.

Social, Behavioral and Economic Sciences

The following must be included in your NSF-SBE data management plan:

• Roles and responsibilities
• Expected data and data formats
• Dissemination plans and policies
• Period of data retention
• Data storage and preservation of access
• Any additional data management requirements (such as institutional polices or best practices)