Anonymisation core concepts

Pseudonymisation or de-identification of data refers to the process of removing or masking direct identifiers. This means that any identifiers that might directly lead to the reidentification of data subjects, such as their names and addresses, are either replaced with pseudonyms (fake names), codes, or redacted. As only direct identifiers are removed, the data remains indirectly identifiable; hence, data protection legislation applies to any additional processing of the data.

Anonymisation refers to the process of ensuring that the risk of identifying a data subject is negligible. This involves more than de-identification and often requires further alteration or masking of data, generally by treating indirect identifiers.  Anonymisation allows data to be shared ethically and legally while preserving confidentiality.

Direct and indirect identifiers

Individuals may be identifiable through:

• Direct identifiers, such as names, full addresses, national identification numbers, email addresses, raw audio or video recordings of identifiable individuals.
• Indirect identifiers, which may not identify someone on their own but could enable identification when combined with other information. Examples include age, education level, occupation, ethnicity, geographic detail, income and rare characteristics.

Effective anonymisation requires considering and modifying both types of identifiers.

Reidentification risk

Disclosure risk depends not only on the data itself but also on the availability of external information that could be linked to it. Factors that influence risk include:

• data size and uniqueness of records
• geographic and temporal detail
• sensitivity of variables
• linkage with other data
• intended access conditions and reuse licences.

Assessing reidentification risk is therefore a contextual judgement that should be revisited as projects evolve and sharing plans become clearer.

Ethical communication through participant communication

Where participant communication materials are used, it plays an important ethical role in establishing transparency and trust between data producers and participants. Materials provided to them helps ensure that participants understand how their data will be used, protected and potentially shared, and allows them to make informed choices about participation.

Participant materials should be aligned with planned anonymisation and sharing strategies. For example, if identifiable elements will be retained this should be communicated clearly. Even where participants agree to data sharing, data producers should continue to minimise disclosure risk and prevent foreseeable harm.