NSF BIO Data Management and Sharing Plan Guidance

NSF recently issued a supplement to update the Proposal and Award Policies and Procedures Guide (PAPPG) to reflect changes for data management and sharing and to align with the Gold Standard Science implementation plan. The supplement describes revisions to these PAPPG sections:

1. Public access to research products (chapter XI.D.2.c).
2. Dissemination and sharing of research results (chapter XI.D.4).
3. Data management and sharing plan (chapter II.D.2.(i).(ii)).

Among these updates is a change to the method for submitting the Data Management Sharing Plan (DMSP) as part of a proposal. The DMSP must now be submitted by completing a form in Research.gov. Submitters will describe up to four data types or research products and provide information for each that is relevant to the following categories: Limitations or exceptions; standards and metadata; provenance; public archiving; timeline for public accessibility; data availability; and accountability. Read additional instructions for completing the DMSP form on Research.gov.

Additional context and guidance specific to the NSF BIO research community is below. Please check individual funding opportunities for any unique requirements for DMSPs.

The BIO Directorate recognizes that data management practices, standards and data life cycles vary across research communities and evolve over time. Accordingly, scientific communities should help shape standards and expectations for data management and sharing. While DMSPs may differ by field, each plan should be appropriate for the research outputs, assets and resources generated and align with current best practices and standards in the proposed research area.

The following overarching expectations apply:

• Per the tenets of reproducibility and transparency of Gold Standard Science, NSF policy states that publications and supporting data are to be made publicly and freely available at the time of publication.
• DMSPs are expected to include the use of recognized, accessible, community-accepted repositories that conform to appropriate national and international standards for such facilities. Where possible, unique, resolvable and persistent identifiers [such as digital object identifiers (DOIs) or uniform resource locators (URLs)] should be assigned. DMSPs relying on self-publication on personal or laboratory websites, instead of using available publicly accessible repositories, may be considered inadequate.
• In recognition of certain research security risks associated with bulk sensitive personal data, including personal health data and human genomic data, the DMSP form provides a drop-down menu to describe data access restrictions. Common restrictions are populated on the menu, with the option to add additional restrictions as appropriate. A justification must be provided for any access limitations.
• DMSPs should be consistent with the FAIR (Findable, Accessible, Interoperable and Reusable) Guiding Principles for scientific data management and stewardship.
• As appropriate, DMSPs may also include best practices regarding the CARE (Collective Benefit, Authority to Control, Responsibility and Ethics) principles for Indigenous data governance.

Costs associated with implementing the DMSP throughout the award period, including personnel time and other expenses required for the management, preservation, access and sharing of research outputs, assets and resources, are allowable. If applicable, these costs should be included in the appropriate budget categories and justified in the budget narrative.

Under "Results from Prior NSF Support" in the proposal, principal investigators (PIs) and co-PIs must report evidence of research outputs, assets and resources generated under a previous or ongoing DMSP of an NSF-funded activity, including availability, accessibility, preservation and sharing. Any research output, asset or resource identified in the "Results from Prior NSF Support" section, for which a unique, resolvable, persistent identifier (e.g., DOI, URL or similar) exists, should also be included in the "References Cited" section using that identifier.

Implementation of the DMSP is considered part of the annual and final project reports. Research outputs, assets and resources made public during the reporting period should be listed under "Other Products" on the annual or final project report forms. They should also be added to the NSF Public Access Repository (NSF PAR) as "Products" using relevant persistent identifiers, prior to the submission of the annual or final project report. Products from the research should also be added to NSF PAR after the end of the project, as appropriate. PAR stores the metadata and a link to the product, but it does not replicate the product or dataset and should not be relied upon as the sole archival mechanism for a project.

For each type of data (e.g., research output, asset or resource) the following information should be included in all project reporting (as applicable):

• The standards, metadata and provenance for each data type or product.
• The publicly accessible repositories or collections where the data or products are (intended to be) stored and from which they may be disseminated or shared. If available, include the unique, resolvable and persistent identifiers (such as DOIs, URLs or similar) for each publicly released product referenced in the report.
• The timeline for the sharing of data or research products.
• The identity of the person(s) responsible for sharing each data type or research product.

Any changes from the original DMSP should be discussed with the managing program director and should be reported in the "Changes/Problems" section of the annual or final project report.