Monday, October 04, 2010

A Few Papers To Keep In Mind When Talking About p-Values

The first paper's title is very provoking: Why Most Published Research Findings Are False.

The second one (The Cult of Statistical Significance) is instructive for anyone who thinks about quoting p-values.

The third one is by the late E.T. Jaynes, one of the overlooked great men of science: Confidence Intervals VS Bayesian Intervals.

Of course, everyone in research should be aware of Peter Norvig's Warning Signs in Experimental Design and Interpretation

No comments: