Metadata describes other data. By including this additional information about your data, its relevance and context can be easily found when needed.
Metadata is not a new concept, it has always been associated with data, but it was never as important as in today's fast moving digital world. Consider your music or film collection: you record the title, authors, release date, producers, directors, etc. without thinking about it. You may also save the artwork, the studio, or the format it was released in such as LP, tape, CD, MD, Video, super 8, DVD, Blu-ray, 3D, etc. All this information is metadata and allows you to make sense of the data and search the collection for the track that you're looking for.
When it comes to research data, you can record additional information to make your data assessible and easily reusable. This could be as simple as ensuring that your tables have header columns and your images have a description.
Changes in Government policy and subsequent changes to research funding requirements have resulted in the necessity to make metadata available alongside research data in open access archives. This preparation of metadata is intended to make all datasets, no matter how complicated, accessible for reuse.
Metadata is a significant and evolving part of current research practice and, as such, has specific descriptive standards such as the Common European Research Information Format (CERIF), the Dublin Core Metadata Initiative (DCMI), and the Data Documentation Initiative (DDI). These standards enable researchers to share, access, interpret and reuse research data.
The purpose of metadata is to make data understandable when the creator is not there to explain it. So you need to include basic information such as: Title, Creator(s), Publisher, Publication Date, and Identifier (DIO or Handle). You may also want to include the subject, date it was created, language, resource type, format, storage location, rights, access information, keywords, version - anything that adds value to the data that will make it reusable. If you are required to publish your data in a particular archive, you should check their requirements for metadata.
Metadata can take the form of embed documentation, supporting documentation, or catalogued metadata. The majority of formats will record information such as creation date, author(s), version; however, MS Office documents also allow additional information to be added to the file's properties including title, organisation, subjects and keywords to name a few.
When you upload your data to an archive, you will likely package it into a single dataset related to the project or to a publication. In this case, you need to include metadata that complies with the archive schema, but also metadata that describes the containing files. The basics are still required (Title, Creator(s), Publisher, Publication Date, and Identifier) but additional information such as subject, language, resource types, formats, storage location, rights, version, alternative identifier, research grant year, a link to the publication, size, and description are optional metadata in the UHRA.