How do you explain the logic behind your choice of metrics and ensure your findings are statistically sound?
