How do you handle a situation where a non-technical stakeholder disagrees with the results of your statistical model?
