DISP=MOD is used ONLY for sequential datasets. It opens the dataset for writing at the end, extending it if it exists, or creating it if it doesn't. It CANNOT be used for partitioned datasets (PDS), as they require different handling. Option D correctly states this restriction.