How do you handle outliers in your data, and how do you justify your decision to include or exclude them?