SocraticGadfly: Why you can't really trust medical studies

April 06, 2011

Why you can't really trust medical studies

From XKCD, the lowdown on medical studies, with their 5 percent p values (compared to natural sciences, with their 0.01 percent values:

Is there a p-value for Ronald Reagan making erroneous statements after eating Jelly Belly jelly beans?

Note Pane 5, Line 5.

No comments: