The term is a catch-all for all types of descriptions of data. Metadata is to inform researchers about different aspects of the research data. To give a few examples:

• The origin, or provenance, of the data. E.g. the specific location of a source in literature or an archive, or a description of the participants and their numbers.
• An overview of the meaning of the headers in an tabular representation of observations, or the contents of columns and the coding used, e.g. F stands for female and M stands for male.
• A description of the naming convention used in a folder tree.
• A description of the contents of a data package in a repository, its retention period, and the license for reuse of this content.