| Consulting Services |
Free Podcasts |
Worked Examples |
Statistical Quality Analysis |
Miscellaneous Statistical Topics |
Design and Analysis of Experiments (DOE) |
About this Website |
View Article - PDF Table 1 (text) Table 2 (text)
Title: Randomization in Designed Experiments
Author: Keith M. Bower
Publication Source: American Society for Quality, Six Sigma Forum, December 2004
Abstract
This article addresses the importance of randomization in designed experiments.
Notes
A key reason for this paper was to hopefully have more practitioners read George Box's referenced paper: "Must We Randomize Our Experiment?" My paper is essentially a trivial mathematical simulation showing how something can go wrong when a non-stationary disturbance is in the "background," and shows how randomization can address such issues. Box's paper provides some excellent advice, and I strongly advise practitioners to read it.
© KeithBower.com LLC. All rights reserved. Disclaimer