What SQL techniques would you use to identify missing data, duplicates, or anomalies before analysis?