FAIR Data
Data Collection and Sharing
What makes research data more interoperable? And what other characteristics encourage data sharing and reuse? First, let’s define some key concepts:
- Variable (Measurement) : The quantification or description of how data is defined (for example, a question on a survey, or an observation such as blood pressure, along with the unit of measurement and other metadata).
- Value (Observation) : What is recorded for the individual response (for example, a number, the degrees in C or F, a selection from a set of response options on a survey).
As an example, consider temperature.
"Core Temperature in degrees F" is a variable, and "97.9 °F" is a value. Sometimes data for multiple variables are collected together using an instrument such as a form (e.g., survey or questionnaire) or measure.
Biomedical research collects data for multiple variables and produces large datasets. In order to share research data, we need to agree on methods that are valid and reliable, and generate results that are reproducible. This is the role of scientists or subject matter experts. Understandably, it’s not the librarian’s role to decide what instruments or variables to use.
But what is the minimum metadata (or data about the data) needed to describe the method, to ensure validity, reliability and reproducibility, and what are the requirements for data description and structure to make the data FAIR? This is where we, the librarians, come in.
(Image Source: iStock Photos, matejmo©)