Accessibility levels at SND
Data described in the SND research data catalogue have different accessibility levels. These levels are displayed with icons that give you direct information about where the data can be found and in what way they are accessible. The accessibility level is determined by the person who describes the data, and this can be done in consultation with SND or the local research data support unit (e.g., DAU).
There are two parts to the accessibility levels:
1. Where can I find the data?
Access to data through SND = Data are accessible in the SND research data catalogue. They can either be downloaded or accessed on request.
Access to data through an external party = Data are accessible through another party. The SND catalogue entry contains information about the accessibility level with the external actor, and a direct reference to where the data can be found.
2. How accessible are the data?
Data are freely accessible = Data are freely accessible by direct download from the website.
Access to data is restricted = Data are accessible on request. Access to data is limited due to restrictions, for example because the data contain personal data or other sensitive information. The research principal may conduct a confidentiality assessment before the data can be released.
Please note: Some data in the research data catalogue have the accessibility level ”Data are accessible by order”. This level is due to be phased out and will be replaced by one of the accessibility levels above.
Accessibility levels in the SND research data catalogue
When you search the SND research data catalogue, the search results will reveal which accessibility level the data material has. If the data are accessible through SND, the entry has an icon with the SND logo. If the data are accessible through an external portal or website, they instead have an arrow. Data that are accessible by direct download are displayed with an open padlock.
Access to data with different types of accessibility
You can access data with direct download through a link in the catalogue entry, where you can generally also find relevant documentation.
If access to the data is restricted, you need to make a request for access to the data. Click the Request data button in the catalogue entry. Access can be restricted, for instance, if data contain personal data or other sensitive information. In that case, the research principal must approve each request before the data can be released. When we need to contact the research principal, it takes a little longer for you to gain access to the data, but we process each request as quickly as possible.
Data that are accessible through an external portal or website can be accessed either by direct download or on request. In the right-hand column of the catalogue entry, you can find information about the level of access from the external portal or website. The entry will also contain information about where you can access the data, e.g., by a direct link to the data material, an URL, and/or contact information.
Choose accessibility for data
In the SND system DORIS, you can describe data that will be made accessible via the SND research data catalogue, as well as data that will be made accessible through an external portal/party. If the data can be found through another portal and you only want to share metadata in the SND catalogue, the data need to have a persistent identifier (PID), such as DOI or Handle.
Data that are shared in the SND research data catalogue can be published with open access or restricted access. The Swedish Research Council recommends ”that research data financed via public funds, and applicable legislation allows to be published, should be published openly on the internet within a reasonable time after the research results have been published”.
If the data need to be protected
If data contain information that needs to be protected due to legal or ethical reasons, access to them must be restricted. This can be data that contain sensitive personal information or other forms of personal data that cannot be made openly accessible. This includes direct and indirect personal data, as well as pseudonymised data. Data can also contain other types of information that is confidential or restricted, for instance localisation data about endangered species, information concerning the safety of the realm, research conducted with private corporations, or copyrighted material. These data are also published with restricted access.
If you want to gain access data with restrictions, you place a request for the data to SND. In this request, you need to enter the purpose for which the data will be used, a project description, and contact details to the requester. The research principal or a representative for the research principal will then decide, possibly after a confidentiality assessment, whether the data material can be released. Note that SND doesn’t make confidentiality assessments for another research principal.
Direct personal data
Data that clearly identify a natural person, such as name, personal identification number, photographs of people, addresses, biometric data, or voice recordings.
Indirect personal data
Data that in combination can be used to uniquely identify a person. They can be detailed socioeconomic information, geographic information such as postal code or place of residence, household members, rare diseases, or treatments.
There are no identifiers in the data material, but there is related information that could connect the data to an individual (for instance a code key, regardless of where it’s located).
Data contain no information that can uniquely identify a person (directly or indirectly), and the code key has been destroyed. Anonymous data are no longer personal data, and GDPR does not apply.
Advice for anonymising data about individuals
Research based on humans almost always contains direct or indirect data that can uniquely identify the research subjects. Before you can share this kind of research data with open access, it is important to review the data to make sure that they don't identify an individual. Below are some steps you can take for a quantitative dataset based on data about individuals:
- Remove all direct identifiers.
- Limit indirect identifiers to those that are necessary for a reproduction of the research, or those that may be reused in new research.
- Aggregate/group data that can contain distinguishing information, such as age, date of birth, income, BMI, and postal code.
- Make it less detailed. You can remove day and month from year of birth, or aggregate data for a higher/broader division, by for instance switching from data on a council level to a regional level.
- Avoid values that stand out by merging the higher or lower values of a continuous variable. You can enter ”80 years and older” or ”30 years or younger”, instead of the actual age if the respective groups become too small.
- Create a coded variable to replace free-form text replies or remove free-form text replies.
- Make sure that no code key has been saved, neither with you nor anywhere else. Remember that you will often need a disposal decision to delete a code key that is kept with a higher education institution.