All the sampling units selected in accordance with Articles 5 and 6 shall be audited either exhaustively or by the application of a subsampling methodology, which shall allow the extrapolation of errors at the level of the sampling unit.
The subsampling methodology shall be based on random selection and may be combined with stratification. In the case of stratification, the audit authority shall randomly select subsampling units from each stratum not exhaustively verified. In the case of PPS and MUS standard approaches, high value subsampling units above the interval of selection shall be audited.
The subsampling method may differ from the method applied for the selection of the main sample.