You are viewing a single comment's thread.

view the rest of the comments →

0
2

[–] tehpatriarchy 0 points 2 points (+2|-0) ago 

On a related topic the book "The Cult of Statistical Significance" is a good read.

Summary of common errors:

  1. If it is statistically significant it is practically significant.

  2. If it is not statistically significant we showed there is no effect.

  3. Effect sizes don't matter, just 'significance'.

The Vioxx case as described in the book will make you think twice about taking any pharmaceutical.