To Randomize or Not to Randomize, That Is the Question – Mini Paper Taken from Winter 1998 Newsletter

Abstract: I use a full factorial design with 3 factors, 2 levels each, to show that randomization does not always alleviate the problem of a time dependent factor that I call “drift” affecting our experiment. I also make some suggestions on how solve this problem.

Keywords: Randomization