Data Management Plan Nsf Example

Standard Operating Procedure: NSF Data Management Plan (DMP) Architecture & Execution

1. Document Control Block

• Document ID: SOP-TR-NSF-DMP-042
• Effective Date: October 24, 2023
• Version: 2.1.0
• Review Cadence: Annual
• Author: Julian Vance, Chief Architect, Template Registry

---

2. Executive Summary & Purpose

This Standard Operating Procedure (SOP) defines the institutional requirements, architectural workflows, and compliance standards for generating, reviewing, and archiving National Science Foundation (NSF) Data Management Plans (DMPs). The objective is to ensure 100% compliance with NSF PAPPG guidelines and institutional mandates, guaranteeing that all digital research outputs are systematically curated, preserved, and made accessible for reuse.

---

3. Scope & Prerequisites

Scope

Applies to all research grants, cooperative agreements, and institutional proposals submitted to the NSF by Template Registry personnel and affiliated research units.

Prerequisites & Required Tools

• Access to the NSF Research.gov portal and FastLane.
• Active institutional subscription to the DMPTool (dmptool.org).
• Secure institutional storage infrastructure (e.g., AWS S3 Glacier, Institutional Research Data Repository).
• Markdown-capable documentation editor for version control.

---

4. Roles & Responsibilities (RACI Matrix)

---

5. Step-by-Step Procedure

Phase 1: Pre-Submission Scoping & Architecture

• Identify all anticipated data types (quantitative, qualitative, physical samples, code, workflows).
• Calculate total anticipated storage volume (in terabytes) over the lifecycle of the award.
• Select appropriate long-term institutional or disciplinary repositories (e.g., Dryad, ICPSR, GitHub Zenodo integration).
• Verify compliance with specific NSF Directorate requirements (e.g., GEO, BIO, ENG data policies).

Phase 2: Drafting the DMP Core Elements

• Products of Research: Define formats (CSV, NetCDF, JSON, PDF/A) and metadata standards (Dublin Core, DDI, Darwin Core).
• Access & Sharing: Establish timeline for data release (typically at publication or within 12 months of award expiration). Detail access restrictions for human-subject data.
• Re-use & Redistribution: Outline licensing frameworks (e.g., CC-BY-4.0 for data, MIT for software code).
• Archiving & Preservation: Specify physical and digital preservation strategies ensuring longevity of at least 10 years post-award.

Phase 3: Review, Validation, & Institutional Sign-Off

• Run the draft through the DMPTool institutional validator to check for missing required headers.
• Submit draft to the Chief Architect for technical metadata verification.
• Route finalized PDF to the Office of Sponsored Programs (OSP) 5 business days prior to NSF deadline.
• Upload final DMP to Research.gov as a supplementary document attachment.

---

6. Quality Assurance & Pro-Tips

Best Practices

• Standardized Naming Conventions: Enforce strict file-naming syntax (YYYYMMDD_ProjectID_DatasetName_vX.X.ext) across all project collaborators.
• Persistent Identifiers (PIDs): Mandate the inclusion of ORCID iDs for all key personnel and DOI minting for all published datasets.

Common Pitfalls

• Pitfall: Vague statements like "Data will be stored on a lab server."
  • Correction: Specify redundant cloud storage with daily snapshots and long-term repository deposition.
• Pitfall: Failing to allocate budget for data management tasks in the proposal budget justification.

Metric Thresholds

• Metadata Completeness: 100% of generated datasets must include a README.txt file adhering to the NSF standard template within 30 days of data collection.
• Repository Ingestion: 95% of primary research data must be accessible via public repository within the stipulated 12-month embargo window.

---

7. Frequently Asked Questions

Q1: What is the required length of an NSF Data Management Plan?

A: The standard NSF DMP must not exceed two (2) pages in length, including all bibliographic references and sub-sections, formatted in standard 1-inch margins and 11pt minimum font size.

Q2: How should proprietary or human-subject data be handled in the DMP?

A: Explicitly state the protocols for de-identification, secure enclaves, and controlled access tiers. Detail how data will be anonymized before deposition or why an exemption/restricted-access repository (such as dbGaP) is utilized.

Q3: Are software and code repositories covered under the standard DMP?

A: Yes. Any custom code, scripts, or computational workflows generated under the award are classified as research products and must include a designated code repository (e.g., GitHub linked to Zenodo) with an open-source license.