One of the primary purposes of BIDS is to preserve enough metadata in the dataset to make the dataset understandable even without access to the responsible researcher and to permit easier re-use of the data. Metadata is essentially data about data. An exhaustive overview of metadata types is beyond the scope of this guide. But in the context of BIDS, we are generally referring to crucial elements that are necessary to understanding data about the study and its data owners, info about the images and how they were acquired, as well as information about the behavioral data.
Examples of metadata included in a BIDS dataset:
- A README file at the first level of all BIDS datasets may include basic information, for example, about the study name, any papers resultant from the data, the data's owners, collaborators, location of the data and study protocols.
- Metadata is also contrained in the file naming conventions themselves. At a glance you can tell what participant, session, type of data (eg. structural MR scan, fMRI or EEG) and task a file pertains to.
- Metadata also includes information about the equipment used (e.g. brand name, number of channels) and any necessary details about how it was used (e.g. scanning sequence used, repetition time, echo time, etc.)
- Metadata is also data about the tasks performed during the session, including study design, content, stimulus onset and duration.
- Lastly metadata refers to data about unexpected situations, eg. if a participant moved frequently during scanning, if an EEG electrode was not functioning properly, etc.