The Book of Why: The New Science of Cause and Effect

by Judea Pearl

But Fisher was right about one point: once you remove causation from statistics, reduction of data is the only thing left.