Being FAIR; Having Trust: How clear uncertainty information can increase
the accurate reuse of our data.
Abstract
Data that are FAIR demonstrate specific characteristics including: ease
of discovery, ability to access, community acceptable formats allowing
interoperability, and information that supports the decision for reuse.
The process used to determine data reuse is commonly called “fit for
purpose” or “fit for use”. These criteria are defined using relevant
factors established by the community for which the data was originally
created, and also a “best effort” for criteria needed by other
research communities. The FAIR Data Principles support robust
documentation of datasets to include the necessary information for
reuse. An important part of that documentation, or metadata, is clear
documentation of the quality and uncertainty related to the data being
considered. When this information is not complete, data has a higher
tendency of being used incorrectly leading to inaccurate research,
rejected papers, or even retracted papers. The importance of data
creators to make their data FAIR – including uncertainty information –
directly improves the transparency and integrity of our science today
and into the future.