Review metadata

DORIS has a minimum set of metadata elements that must be provided for a data description to be published. These requirements ensure that research data can be found, accessed, and reused, and that the metadata can be disseminated through other systems. Meeting these minimum requirements is an important step towards fulfilling the FAIR principles.

Fields marked with an orange-red symbol (·) in DORIS are mandatory and must be completed before a data description can be published. The required fields vary depending on the subject profile selected in the form. Some fields are mandatory for all profiles. Non-mandatory fields should still be completed as fully as possible.

Mandatory metadata fields must be provided in both Swedish and English. In fields where terms are selected from a controlled vocabulary, translation is handled automatically. However, in free-text fields, you must manually add a translation in the form. The translations do not need to be exact, but the same level of information should be available in both languages. For longer text entries, it may be appropriate to refer to the catalogue entry in the other language, but an introduction should be present in both languages.

Researchers may submit a data description even if some mandatory fields are missing. In such cases, the reviewer is expected to help provide the missing information. It is technically possible to publish a data description without completing all mandatory fields, but this is not recommended. Such publication should be reserved for exceptional cases; for example, when the PID requirement can be waived (see below under What applies to data descriptions that only contain metadata?). A note explaining the missing information should be added in the Notes section in DORIS.

Good metadata strike a balance between general information and detailed content. The Description field is especially important. It should explain the context and nature of the data in a way that is understandable even to researchers outside the field. It should also provide enough detail to allow potential users to quickly assess whether the dataset is relevant and worth exploring further. Other metadata fields, such as those under Collection and method, are typically aimed at a more specialized audience.

Some metadata fields contain links to external sources such as publications, websites, or related resources. These links should, whenever possible, be persistent identifiers such as a DOI, handle, or URN. If no persistent identifier is available, a standard URL can often be used, but since URLs do not guarantee long-term accessibility, they may need to be verified periodically.

Make sure the metadata are reasonable and useful, that no fields contain meaningless values, and that there are no spelling errors or typos.

DORIS can be used to share metadata for datasets that have already been published elsewhere. This is done by ticking the box I only want to share metadata. It can be more challenging to review these types of data descriptions, as you may not have access to the dataset (it may be openly accessible through another portal, or it may require you to request to access the data).

Note: This checkbox should not be used for datasets that cannot be shared openly – such access restrictions should instead be indicated using the Data access level field.

Data descriptions that only contain metadata are not assigned a DOI in DORIS, as the description does not include a dataset. Instead, an existing persistent identifier (PID) for the dataset must be entered in DORIS, in line with SND’s Policy for Persistent Identifiers (PID) on Research Data. Exceptions to the PID requirement may be made when the data description refers and links to a resource collection (such as a database, catalogue, or portal) that itself contains multiple datasets.

Ensure that the licence and copyright information do not conflict with the selected data access level, your institution’s local policies, or other applicable frameworks. As an example, datasets with restricted access should not have a Creative Commons licence applied. 

It is crucial that the information provided here is correct. A detailed review and discussion with the researcher are often needed to verify this. Datasets that contain personal data may only be shared via DORIS if the researcher’s organization has an agreement with SND. This may involve local storage or a data processing agreement with SND.  

Also indicate what type of personal data are included. This is important for both prospective secondary users and internal processes for assessing requests for access and disclosure of the data.

If it is not possible to fully determine whether personal data are included – for example, due to the size or complexity of the material – we recommend selecting Data contain personal data and using the Type of personal data field to describe any potential risks related to the personal data.

Check that the selected data access level aligns with the content of the data – particularly regarding personal or protected information. If the data contain information that cannot be shared openly (e.g., sensitive personal data), the access level should be set to Access to data is restricted, which means a request to access the data is required before access is granted.

Read more about data that contain personal data on GDPR and personal information.

Remember that even if data cannot be made directly downloadable, documentation files should be. This is important so that a prospective data requester can obtain as much information as possible to decide whether the material may be useful to them.

Titles should ideally be provided in both Swedish and English, but this field is exempt from the general bilingual requirement as a suitable Swedish title may not always be available for the purpose of data publication.  

The Alternative title field can be used to enhance discoverability, for example by providing a title in a third language.

Avoid using the same title for the data description as any publication associated with the research results. In such cases, use a title such as “Data for: [Title of the article]”.

This may be one or more individuals or organizations. It is relatively uncommon for both individuals and organizations to be listed as creators of a dataset. An organization may be listed as the creator when the dataset stems from a large research project or collaboration. For example, the SOM Institute lists the organization as the creator of its surveys. 

The names entered here will be included in the data citation. Individuals or organizations that contributed but should not be included in the citation should instead be listed in the Contributors field (see next section, Administrative information). Whenever possible, include ORCID IDs for individuals and ROR IDs for organizations to improve interoperability and findability.

If a researcher’s institution does not have an official Swedish name, use the English name in both language fields rather than leaving the Swedish field blank. 

This is a central field in any data description. Description should provide users with a clear understanding of what kind of data are involved and the context in which they were collected or generated.

A good description balances general information – accessible to non-specialists – with sufficient detail to enable researchers with domain knowledge to assess whether the dataset may be relevant to them for further research. Larger and more complex datasets from completed projects may benefit more from broader descriptions aimed at a wider audience, than what smaller, article-specific datasets do. For highly specialized material, it may be acceptable to provide a more targeted, specialized description.

Description is a mandatory field in both Swedish and English. However, the text does not have to be identical; a shorter summary in Swedish with a reference to the English catalogue page for more information is acceptable.

If you draft a simple description yourself, make sure the researcher reviews and approves it. Consider including a reference in the text to the English page for further information. Always ask the researcher to verify what you have written, especially if the research field is unfamiliar to you.

Research principal refers to the organization within which the research was conducted and which bears ultimate responsibility for it. If there are multiple research principals, enter the organization responsible for making the data accessible. You can, for example, ask the researcher which organization is responsible for archiving the data or which organization submitted the ethics application. 

This field can be used to list additional research principals, e.g., in collaborative projects. In collaborations, such as those between a university and a regional authority, it may be unclear which organization should be listed as research principal. The researcher should usually know this, but you should be alert to cases where the creator and principal investigator come from different organizations, or where the metadata indicate a collaborative project. If in doubt, ask the researcher for clarification.

Keywords improve the findability of data. A data description must include at least one keyword, but the more keywords provided, the more findable the dataset becomes. Whenever possible, keywords should be selected from one of the controlled vocabularies. If a suitable term is not available, free-text keywords can be entered. The advanced keyword search allows you to explore terms within specific subject areas and vocabularies. You can also search using the English version of the keyword lists, as some terms do not have Swedish equivalents. 

As a reviewer, you are encouraged to help the researcher by suggesting or adding more relevant keywords.