From: Stan Hilliard
Date: 6/13/2005
Time: 11:47:01 AM
Hi Connor04,
I think that customer order quantity is not the best basis for forming lots for the purpose of acceptance sampling. It is better to base the accept/reject lot on process rational subgroups.
Try to define lots in such a way that they are as homogeneous as possible. When an important variable changes that can effect quality -- that should define the beginning of a new lot.
Another aspect of lot size to consider is in connection with the amount/value of product involved. Hugh lots can create a crisis if one has to be rejected.
There are many ways to make sampling cost effective and efficient. Especially, be wary of AQL based Mil-Standard plans, and of indiscriminately using C=0 plans. Here is a link to other "best practices" that apply.
samplingplans.com/modern3.htm
I have a few questions about your application. Is the Sampling Accumulation Method a formal sampling procedure?
What kind of product do you manufacture?
Are you dealing with measured variables or attributes of both?