Research Data Management
Research Data Management (RDM) describes the collection, organization, storage, preservation, and sharing of data - whether quantitative or qualitative - used in a research project. In other words, RDM is a planning and implementation process that spans the research data lifecycle in order to ensure effective and responsible data-driven research.
Detailed information on RDM can be found at the following links:
- Research Data Management @ Pitt - library guide from the University of Pittsburgh covering the RDM lifecycle, including writing a data management plan, rights and permissions, and finding, organizing, describing, storing, and sharing data.
- Keeping Up with Research Data Management from the Association of College & Research Libraries
RDM @ Harvard
Many resources exist across Harvard to help you with RDM in your projects.
- Harvard Library's Research Data Management Program (HLRDM) advises researchers and connects you to resources spanning the research data lifecycle. The program offers online educational tools, such as video tutorials, case studies, an electronic lab notebook guide, background literature, and slides from past presentations and classes.
- HLRDM also periodically offers in-person workshops on topics like data visualization and writing a data management plan. View and register for upcoming workshops here.
- Research Data Management @ Harvard provides information and resources to help you manage research data through all stages of the data lifecycle:
- Planning Data Management
- Data Acquisition and Collection
- Storage, Security and Analysis
- Dissemination and Preservation
- If you are working on grant- or contract-funded research, you will likely have to create and/or maintain a data management plan. Harvard affiliates have access to the DMPTool, a widely used data management plan creation tool that can help ensure that you meet your funder's requirements.
A critical issue in RDM is data security, particularly if the research produced depends in whole or in part on confidential and/or sensitive data. Though measures to ensure data security tend to be most visible during storage, they are important to keep in mind at all stages in the RDM lifecycle.
At Harvard, data security is governed by the University-wide Harvard Research Data Security Policy (HRDSP). It follows the Policy Statements of the Harvard Information Security Policy while providing additional guidance specific to RDM. The HRDSP describes the roles and responsibilities of all researchers at Harvard, and outlines protocol around the handling of particular types of data. Researchers seeking support on compliance with the updated HRDSP - rolled out on July 15, 2020 - should consult the Research Data Safety Knet page (Harvard Key login).
Fundamental to the HRDSP is the classification of data into different risk levels, reflecting the basic principle that more exacting security requirements must be implemented as the risk associated with the research data increases. At Harvard, there are five levels of data security classification:
- Level 1: Non-confidential research information
- Level 2: Benign information to be held confidentially
- Level 3: Sensitive or confidential information
- Level 4: Very sensitive information
- Level 5: Extremely sensitive information
These levels are listed in further detail in the Harvard Data Classification Table, and the handling requirements associated with each of the levels are described at this link. Harvard's facilities for appropriately storing data of these varying security levels are listed here.
Additional resources related to data security at Harvard include: