A dataset contains a few extremely large salary values. The analyst wants to describe the typical central value without those extremes dominating it. Which statistic is most suitable?

1. The number of distinct column names
2. The sum
3. The maximum
4. The median