A summary of the evidence that most published research is false | Simply Statistics

This figure relates probability for and agains...
This figure relates probability for and against a true-false assertion, and/or the associated odds ratio, to surprisals against and for (as well as their difference, i.e. the evidence) in bits. It may be used to plot data on risk and evidence using these log-probability measures. (Photo credit: Wikipedia)

A ‘must-read’ piece on two major ideas wandering about the literature at the moment, namely that:

via A summary of the evidence that most published research is false.

Author: Shane O'Mara

Neuroscientist