About Research Data Management

Research Data

Generally speaking, research data are the observations or evidence you analyze to produce research findings and can take any form including text, measurements, audio/video recordings, or any other bits of information you use in your analysis. Everyone who conducts research, in any field, has research data.

In academia, our concept of research data is heavily influenced by funding sources in the federal government who define research data as “digital recorded factual material commonly accepted in the scientific community as necessary to validate research findings.” This federal definition is focused on the data records you are encouraged to share at the end of a research project and includes any digital data you analyze to produce findings. This generally does not include other important records like lab notebooks, preliminary analyses, drafts of papers, or communications with colleagues, which federal funders usually do not expect you to share.

Research Data Management

As the proof underlying your research findings, your research data need to be properly managed so they are organized and understandable, both to you and to others in your field who could validate your work. That management starts with planning for a research project and continues all the way through final publication and archiving.

Common data management activities include:

• Writing a data management plan
• Planning for data collection
• Organizing and naming your files
• Storing and backing up data
• Writing documentation like a ReadMe file or codebook
• Cleaning data for analysis
• Publishing and archiving data

FAIR Data

The FAIR Data Principles are widely accepted data management guidelines for making data:

• Findable
• Accessible
• Interoperable
• Reusable

The FAIR principles ensure data can be understood, used, and cited, which helps to maintain research integrity and increase research impact. To make your data FAIR, aim to use standard terms and identifiers whenever possible, start documenting your data early, choose file formats that are compatible with multiple software tools, and publish non-sensitive data in a repository with a Digital Object Identifier (DOI) and a license permitting reuse.