Statistics Done Wrong: The Woefully Complete Guide

by Alex Reinhart

These rules are often violated in the neuroimaging literature, perhaps as much as 40% of the time, causing inflated correlations and false positives.