Describe a time when you found a critical discrepancy in a dataset right before a major presentation. How did you handle it?
