When research data are going to be stored long term, they should have a detailed description by metadata that follows the current standards in the research field. A good description by metadata is also a central part of the FAIR principles. The descriptions by metadata should be openly available, even if access to research data is limited. In addition, the metadata should remain, even if the data is destroyed.
Metadata describe the most important features
Metadata is information about another resource, which in this case is research data. They make it possible to find and understand the resource without having to have a human user or a machine to manage the resource.
Metadata has a value of its own and should always be retained. A description by metadata makes it possible to understand the context that underlies the publications of a project and to use similar methods to conduct similar studies, even if the research data is not accessible or no longer exists. Long-term storage of metadata is often also considerably less expensive than storing large amounts of research data.
Comprehensive, high-quality metadata descriptions are essential to meeting the FAIR principles. The FAIR principles do not require that research data are open, but if access to data is limited in any way, this should be stated in the metadata description.
When a research project is finished and its data are to be stored in an archive or repository, it is time to make a detailed metadata description of the project data. The metadata description is made according to the standards and using the controlled keywords that are relevant in the research area.
This work is facilitated, if you already at the beginning of the project think about which metadata standard you likely are going to use. The metadata standard tells you what information will be needed to describe your data, and, hence, you can collect, document and compile reference material for metadata continuously.
A metadata description contains different types of metadata, which together make it possible to understand a resource without having direct access to it. Metadata should, as far as possible, be readable both by humans and by machines. Then the resource becomes searchable through as many channels as possible. The more information the metadata provide, the better it is.
The structure of how metadata is presented may differ depending on the metadata standard used. Use a metadata standard that is established in your research field and fits your data and publications.
A detailed metadata description may contain information on:
The content and structure of the resource
Data describing the content of the resource can, for example, be:
Summary or abstract
Table of contents
Relationships that link materials and data that are based on one another, such as code keys and databases
The origin of the resource
Data describing the origin of a resource can, for example, be information on:
The generation of the resource
Data describing the generation of the resource can, for example, be information on:
Context, or factors that may have affected the outcome
The time when data was created and last updated
Other data sources used
Other data sources may, for example, consist of re-used data from own or other research projects or of data from public data sources. If data from other sources than the research project have been used, metadata should include information on this and provide a detailed reference to these data sources.
It is important to describe the context between different data and their metadata. For example:
Which code keys apply to which documents?
Which documents describe units and categories that apply to measurements?
How are different versions of question sets related to each other?
How are different ways of retrieving information from informants related to each other?
Which data has been used to enrich other data?
Which recordings belong to which transcripts?
Administrative information, access and restrictions
Administrative and technical metadata contain information about access to the resource, such as:
When was metadata created and last updated?
What rights and licenses do apply?
Is allowed to read and disseminate the data?
Is it allowed to reuse the data or is a permission needed?
Is the material accessible and how can access be obtained?
Are special permits needed and how does one apply for it?
Who is responsible for storing data?
Who is available as a contact and what are the contact details?
What resources for technical access are available?
Are there persistent hyperlinks?
Are there technical protocols for access?
Are the resources placed in a physical room?
Register research data at the Swedish National Data Services (SND)
The University uses the national catalog for research data of the Swedish National Data Services. The research data team of the library reviews the registrations that our researchers submit through the SND service Doris to ensure that the registration meets the requirements for registration and comply with the FAIR-principles. Log in to "My pages" at SND to register research data.
As a researcher you can get help from both the library and from SND, if you have questions about the registration process and metadata.