As usual, an excellent post by Ben Goldacre on the need for evidence based social policy. We all know that RCTs are difficult to set up in many social policy contexts, but difficult is not the same as impossible. Often it just requires that politicians have the balls to say that unless you want to sit around waiting for an instrument to be provided by the Gods of Natural Experiments the choice is either ignorance or randomization. If you don't like the latter then you have to believe that the cost of ignorance is smaller than the cost of knowledge. It's a perfectly consistent position but I have no idea how anyone could claim to know that the status quo is superior to an alternative without ascertaining the facts of the matter.
Has Statistics become corrupted? Philip Stark’s questions (and some
questions about them)
-
In this post, I consider the questions posed for my (October 9) Neyman
Seminar by Philip Stark, Distinguished Professor Statistics at UC Berkeley.
We didn’...
3 hours ago
No comments:
Post a Comment