If you were given a dataset with significant discrepancies, how would you approach identifying the issue?
