When reading from a Hashed File Stage in DataStage, column values can be re-ordered by specifying their position in the metadata definition. Unlike database stages where column order follows the source schema, hashed files allow explicit positional mapping.