Anonymization is a process that completely removes the possibility of identifying participants. After anonymization, it is impossible to determine to whom the data refers.
Pseudonymization, by contrast, is a reversible process. Typically, a numeric or alphanumeric code is assigned to each participant, and a key linking the codes to the identifying data is stored separately. The pseudonymized dataset must not contain identifying data, but the identities of the participants can be easily recovered by combining the experimental dataset with the key file, which must be stored under higher security measures in order to reduce the risks associated with processing the pseudonymized dataset.
For more information, click here.