[ | Next | Previous | Up ]

Re: skip lot and two point sampling

From: Stan Hilliard
Date: 03 Jul 2002
Time: 10:01:26

Comments

Greetings Bart,

I do not have a software tool for lot skipping plans at this time. Actually, I have written a program for CSP-1, CSP-2, and CSP-3, but it will not be ready to sell soon because I am in the process of updating the current two-point programs now.

The best that I can suggest for now is to use a two-point plan to sample the lots that are not skipped, and to use a CSP plan to determine which lots to skip. The CSP plans apply to determining each individual unit, item, etc. tested to be defective or not. To use it for lot skipping, you define a lot as a unit. A lot that is accepted by the two-point plan becomes a non-defective unit for the purposes of the CSP (skip-lot) decision.

The CSP plans in the literature have oc-curves associated with them. The percent defective on the x-axis of those oc curves is interpreted as the percent of lots that are rejectable by the lot plan.

By using a two-point plan for lot decisions and a CSP plan for lot-skipping decisions, you have two oc-curves - one curve for each aspect of the scheme. The two curves need to be examined separately. That is, I do not know of any way to combine them into one curve.


Last changed: November 20, 2007