From: Stan Hilliard shilliard@samplingplans.com
Date: 14 Aug 2003
Time: 10:12:10
Greetings Gabriel,
I took extra time to answer your message because I thought that it was quite thoughtful and I wanted to answer it in kind.
YOU SAID: "Stan, I have read about all the previous messages on this subject (C=0), and found you to be pretty contrary to them."
MY REPLY: I do not have any problem with using acceptance numbers of zero. The thing that I have warned about (in other messages) is using C=0 plans UNIVERSALLY for product acceptance -- as a policy, for all products, quality charactistics and types of applications. Some companies have such policies believing that it is good practice.
For product disposition purposes, I think that C=0 plans are appropriate for those characteristics where a defective is life threatening or otherwise call for rejecting the lot. Such plans are designed by either choosing C=0 directly or by using the oc curve to get a lowest practical RQL. In my experience, the majority of measured quality characteristics are not that critical. The traditional scheme of classifying defects as critical-major-minor recognizes this.