If paired analysis is required on the second level with repeated measures, there is a required order for the repeated measures, namely, the order of subjects must be the same for all groups in Factor A must be the same. This is either not documented, or poorly documented.