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