Storage of ongoing research

Yellow data may be stored on a local disk (PC/Mac managed by OsloMet), as well as on a USB flash drive or an external hard drive that is encrypted.

In addition, there is the option of storage on a local disk in tablets and mobile phones configured in the Hub/Intelligent Hub app through OsloMet.

Red data may only be stored on a local disk if the data or the machine is encrypted, and the folder is not configured to synchronize with cloud services.

By cloud solutions, we mean websites and other services where the actual storage takes place outside your own machine and is accessible via, for example, a phone or PC/Mac.

Yellow data may be stored in Microsoft 365, Teams, and OneDrive/SharePoint provided the information is access-controlled and protected with sign-in. Yellow data should be stored with the classification “Confidential” in Microsoft 365 where you are asked to define this. If you are unsure how to set up access control and protected sign-in, contact the research data management team (oslomet.pureservice.com).

It is not possible to store red data in regular cloud solutions.

Backup

OsloMet backs up your files for up to 90 days after they have been removed from the OneDrive (SharePoint) recycle bin. Note that if entire Teams/SharePoint sites are deleted, the backup retention is only 30 days.

Depending on size, classification, and type of research, there may be a need for specific solutions. Below is an overview of what OsloMet offers.

If you are unsure, please contact the research data management team for guidance (oslomet.pureservice.com) on what might suit you and your project.

Note that only TSD is approved for storing directly identifiable health information.

Research infrastructure

(1) Mime’s limitations for red data: Less sensitive red data and/or indirect red data that would be difficult for others to identify or combine may be considered for storage in Mime. The assessment must be based on data protection principles and the security relevant to the specific project, as well as where and how the code key (linking key) should be stored.
 
(2) EduCloud’s limitations for red data: It is possible to use EduCloud for red data, but it is assumed that a risk assessment is carried out in the project with regard to data leakage and data protection impact, since the service has access to the internet and connected devices, and that good access control and workflow routines are in place. EduCloud is offered only for research projects with extraordinary needs.
 
(3) Additional needs: Additional costs may apply, depending on the project, and must be clarified in advance.

If you are collecting information from participants via surveys, the information will also be stored in the technical solution you use. Below you will find an overview of the types of data you can store in the various collection services.

Services for datacapturing

(4) For students and others using personal equipment: As a general rule, yellow data should not be downloaded or synchronized to personal equipment. If personal equipment is used, questionnaires and files must only be opened through the Nettskjema portal (nettskjema.no). If analysis tools such as NVivo, SPSS, etc. are to be used for further processing, a class set in EduCloud will be a suitable platform.

(5) For red data collected in Nettskjema: These must be deleted within a reasonable time (90 days) or moved to an approved storage location such as EduCloud or TSD. Nettskjema is also directly connected to one of these platforms.
For students and others using personal equipment, a voice recorder app is approved for small amounts of red data in student assignments with low data protection impact. Further processing of the data is done in a class set either in EduCloud or TSD.

(6) The agreement with Easyfact is offered only at NIBR.

Both copyright and data protection may apply here, depending on the type of research material you have. The general rule is that yellow, red, and black research data must not be entered into AI tools or tools with AI components as input. This is because those tools are connected to models hosted externally, and not locally with us.

Note that there is a difference between the use of language models in research or research software with AI components versus generic AI tools. For using language models in research, OsloMet offers an AI infrastructure as a technical platform. Here it is also possible to connect models installed locally to software running in this infrastructure. Information about AI in research.

For the TSD-Autotekst software, it is only permitted to enter personal data about others if you have the consent of the person(s) concerned  or have another lawful basis for doing your research.