Static masking involves taking a data set and replacing sensitive fields and values with non-sensitive or garbage data. This is done to enable testing of an application against data that resembles production data, both in size and format, but without containing anything sensitive. Dynamic masking involves the live and transactional masking of data while an application is using it. Tokenized would refer to tokenization, which is the replacing of sensitive data with a key value that can later be matched back to the original value, and although it could be used as part of the production of test data, it does not refer to the overall process. Replicated is provided as an erroneous answer, as replicated data would be identical in value and would not accomplish the production of a test set.