I have dropdowns of employee names. The source is an external CSV, and I use several calculates with pulldata() to pull employee metadate (e.g., email) based on selection.
Sometimes, when going through the Sent folder and choosing Copy data to a new survey, these dropdowns change from the label column (employee name) to the name column (employee number). When this happens, S123 ignores data valedation, and allows the survey to be sent with bad data in the dropdown field (see below).