
Think you know stats? Stuart Hurlbert first described pseudoreplication—a common but serious statistical error—in 1984. Despite widespread knowledge of the error, pseudoreplication is often misinterpreted, and literature surveys show that the error is on the rise in certain fields. Listen to Hurlbert define pseudoreplication and other related errors, plus hear why we shouldn’t dichotomize results as “significant” and “non-significant,” what’s missing from basics stats courses, and what’s next on his list.
Access the Learned Discourse by Hurlbert and Lombardi in the January 2016 issue of IEAM.