Statistical Modeling, Causal Inference, and Social Science
The problems with p-values are not just with p-values: My comments on the recent ASA statement
Andrew Gelman | Professor of Statistics and Political Science and Director of the Applied Statistics Center, Columbia University
1 comment:
This is a brilliant comment. And it applies to climate change research as well. In fact, it stabs at the heart of it, and the more responsible climate journals are now (this year) requiring that papers include proof of their statistics models and how they arrived at them. Some journals are now requiring that they get sign-off from statistics experts that the models are properly constructed.
it’s significance, however, will go over most people’s heads. And Joe Schmoe doesn’t understand this. Neither do the heavy-breathing non-scientifically-trained climate change advocates who pass for brains in our major papers who wouldn’t know a p-value from a piss.
if you want the bottom line, Gelman is saying, warning, that you can make statistics produce any conclusion you want with your data, something that real statistics experts have been wailing for years to deaf ears. Keynes, btw, it recognized as a “statistics GOD."
Post a Comment