|Design and Analysis
of Experiments (DOE)
Title: Why Divide By n-1?
Author: Keith M. Bower
Publication Source: American Society for Quality, Six Sigma Forum, February 2005
From a practical perspective, even if 100% inspection occurs on a process, using n-1 to compute the sample standard deviation is appropriate when performing an analytic study.
This paper had the hidden aim of examining the different types of statistical studies and their relationships to various widely used statistical procedures. Though the "n-1" discussion itself may be of help to practitioners, it was to illustrate analytic studies (in particular) that was my primary goal.
© KeithBower.com LLC. All rights reserved. Disclaimer