Sensitive data

No, if not anonymized and/or no consent for publication resp. archiving is given.

No, provided it is information that directly identifies individuals (e.g. tables with keys used to de-identify a patient by coding / pseudonymization).

Yes, provided irreversible anonymization and consent of data subjects for publication resp. archiving has been obtained. Please however note the following caveats and general conditions:

• In general, for sharing or archiving health-related data, we recommend using a repository specialized in the preservation of such data. The Research Collection can accommodate such data under the conditions stated here, but does not specialize in handling health-related personal data.
• Due to the rapid progress of technical methods for de-anonymization, it is controversial whether clinical research data can be permanently and irreversibly anonymized at all. The Swiss Clinical Trial Organisation writes in its 2022 Guidance: [...] clinical research data should rarely, if ever, be treated and shared as anonymized data even after appropriate processing. Researchers should carefully weigh the benefits and risks and, if in doubt, should consider their data identifying" (p. 22). We recommend an appropriate risk assessment before publication.
• Due to the risk of future de-anonymisation, which cannot be completely ruled out, we generally advise against granting Creative Commons licences for data publications that contain health-related personal data. This gives you greater scope to restrict the accessibility of the dataset at a later date, should your risk assessment change in the future.
• For research projects that involve research with human subjects or the collection of personal data or the use of non-anonymized personal data, an application must be submitted to the ETH Zurich Ethics Committee.
• Please refer to the factsheet Data Protection in Research Projects of the Legal Service of ETH Zurich.

• Data about an identified or identifiable individual that relates to his or her health or disease
• Genetic data

No

To determine whether your research data is highly sensitive personal data that must be classified as ‘strictly confidential’, you must perform a classification and risk assessment in accordance with the Directive "Inventarisierung und Klassifizierung von Informationen an der ETH Zürich" (Inventory and Classification of Information at ETH Zurich).

• Personal data that, if improperly disclosed or processed, would have a serious impact on the health and economic situation of a person (life and limb).
• Depending on the geographical and political context, statements about political or religious views, for example, must be classified as highly sensitive.

Yes, provided

• that data has been classified as non-highly sensitive and non-health-related and
• that the consent of the data subjects for publication resp. archiving has been obtained

Not classified as highly sensitive or health-related

• completed questionnaires, transcripts or recording of interviews that can be directly attributed to individuals
• photographs of identifiable persons
• photographs of private spaces (e.g. houses, gardens) or objects (e.g. license plates of cars) that can be assigned to a person without much effort

Non-personal, strictly confidential research data

No

To determine whether your research data is strictly confidential, you must carry out a classification and risk assessment in accordance with the Directive "Inventarisierung und Klassifizierung von Informationen an der ETH Zürich" (Inventory and Classification of Information at ETH Zurich).

• Unveröffentlichte Forschungsergebnisse, die bei vorzeitiger Offenlegung hohen oder sehr hohen Schaden anrichten können
• Vertraglich als streng vertraulich vereinbarte Forschungsdaten (z.B. mit Kooperationspartnern, Dritten)

No

Information on classification can be found on the export control web pages.

• Software or technologies for military goods
• Software or technologies for dual use goods

Usually no

Such documents are generally not intended for publication or sharing with other researchers, so a public repository is not the most appropriate tool for preservation. We recommend that such documents be archived in an appropriate internal repository of the research group or department.

Data documenting administrative processes related to a research project, e.g.

• Contracts
• Research plans
• Accounting records
• Invoices
• Documents from personnel administration