Abstract

Optimizely is moving away from traditional, fixed horizon hypothesis testing to sequential testing and replacing Type I error control with false discovery rate (FDR) control. This will enable users to confidently check experiments as often as they like, not need to know an effect size in advance, and test as many variations and goals as desired without worrying about hidden sources of error. Read on to understand why, how, and what new tradeoffs exist.

Zum Artikel
Jan
Jan
Autor
Psychologe, Web Analyst, Science Junkie, Star Wars Missionar und Gründer des Blogs

Hinterlasse einen Kommentar

Please enter your comment!
Please enter your name here