33. Outliers and robust analysis

Covers outlier detection, heavy-tailed data, robust summaries, winsorization, and sensitivity checks. Learners decide when an unusual value is an error, a rare event, or the most important part of the story.