Mill’s methods of causal inference (student stuff) .[embedded content] As we all know, R. A. Fisher was not too happy about Mill’s method of difference, since it, according to him, built on the impossible requirement of being able to compare identical units under different circumstances. Fisher instead favoured the experimental method of randomized treatment assignment. But if you cannot assign treatment randomly — as in most observational studies — there is much that supports Mill’s method of difference for making causal inferences.
Topics:
Lars Pålsson Syll considers the following as important: Theory of Science & Methodology
This could be interesting, too:
Lars Pålsson Syll writes The wisdom of crowds
Lars Pålsson Syll writes Three pages — all it takes to change science forever
Lars Pålsson Syll writes Necessary and sufficient (student stuff)
Lars Pålsson Syll writes Science — the need for causal explanation
Mill’s methods of causal inference (student stuff)
.
As we all know, R. A. Fisher was not too happy about Mill’s method of difference, since it, according to him, built on the impossible requirement of being able to compare identical units under different circumstances. Fisher instead favoured the experimental method of randomized treatment assignment. But if you cannot assign treatment randomly — as in most observational studies — there is much that supports Mill’s method of difference for making causal inferences.