Guides: Research Data Management: Data Lifecycle

Data Lifecycle Stages

Research Data Management is a continuum of practices. It continues throughout the course of a research project. You will likely jump around and move between phases in the lifecycle, but you should always start at the Plan & Design phase. During the Plan & Design phase, you will need to know:

your research project,
research stakeholders,
roles and responsibilities,
funder requirements,
data goals,
and challenges.

Use a checklist to help plan and design your work:

Research Data Management Checklist (LMA Research Data Management Working Group)
Start and End Checklists (MIT Libraries)
Roles and Responsibilities Checklist (COGR) (Excel spreadsheet)

Review the Support Your Data Project at University of California https://researchdata.org/ which presents a holistic look at data management best practices. The project provides a framework for thinking about data throughout the project lifecycle.

You may need to create a Data Management Plan during this phase. Additional guidance for creating a plan is available in this guide under the For Researchers tab.

Before launching a research project, design a model for capturing, storing, and organizing your data. Consider project

workflows (record data procedures, workflows, protocols, and responsibilities),
data types (document what types of data will be produced in the project),
metadata standards (use common standards including Common Data Elements or other FAIR metadata standards),
formats (chose non-proprietary formats when available or convert to open, non-proprietary, widely used formats for sharing),
volume (understand how much data will be created as part of the project),
access (document any specialized tools required).

Design how you will store your data:

Data Management general guidance for working with data (DMPTool) https://dmptool.org/general_guidance
File naming tips (Cornell) https://data.research.cornell.edu/content/file-management
File naming and Folder Organization tips (Johns Hopkins) https://dataservices.library.jhu.edu/file-naming/
Build your Directory Structure (Harvard) https://datamanagement.hms.harvard.edu/collect/directory-structure
Write a ReadMe file (Cornell) https://data.research.cornell.edu/content/readme

Store & Manage is a key component of the Data Lifecycle touching on every stage. Researchers will need to plan for:

how data sets will be stored during active working phases (including backups) and for long-term retention,
retention polices set by funder or institution,
data security,
and related costs.

Consider data storage requirements for the project.

Cornell Data Storage Finder tool provides general guidance for selecting storage and collaboration options.
Contact your campus data liaisons to determine computing and storage options available at your institution.
Consider tools that support versioning and collaboration such as GitHub, wikis, and shared drives.

Follow required retention and preservation requirements as established by your institution or funding agency. The Data Curation Network has developed extensive guidance on working with and keeping research data.

CURATE(D) Steps and Checklists (Data Curation Network)
Data Curation resources (Data Curation Network)

Find a repository for sharing and publishing:

NIH Open Domain-Specific Data Sharing Repositories https://www.nlm.nih.gov/NIHbmic/domain_specific_repositories.html
Registry for Research Data Repositories https://www.re3data.org/
Check for your own institutional data repository
Cornell tool for determining data storage service https://finder.research.cornell.edu/storage
Generalist hosting repositories (GREI list from NIH) allow independent submissions or may be part of journal publishing requirements:
SUNY Dryad Instance: SUNY OLIS and some SUNY Campuses are members of Dryad and will be experimenting with Dryad as an open data publishing platform and community.
SUNY SOAR https://soar.suny.edu/ is an institutional repository, but not designed for data storage and sharing.

Data publishing repositories should follow FAIR principles https://www.go-fair.org/fair-principles/

In general, raw data are considered facts and cannot be copyrighted. Community norms for data attribution and scholarly communication are often more successful in documenting origins of data than licensing restrictions when possible.

Data license considerations include the following:

Dryad only allows publishing under the Creative Commons Zero (CC0) for data.
Some data license options using another repository such as Figshare.

Data Lifecycle

Data Lifecycle
Image and Biomedical Data Lifecycle created by LMA Research Data Management Working Group at Harvard Medical School licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

The Data Lifecycle (link to interactive lifecycle) is a representation of stages that occur in your research in regards to how data is collected, used, and stored.