Walk me through a complex statistical model you have built. How did you validate its assumptions and handle outliers?
