From: Stan Hilliard
Date: 29 Apr 1999
Time: 20:53:46
Greetings Mark, I waited a few days to reply hoping that someone else would add a not, but no-one did. My answer is yes, all attribute acceptance sampling plans have oc-curves that go through the point X=0, Y=1.0. So if you define AQL as zero, the producers risk (alpha) will be zero. Refer to the oc-curve graph at
http://www.samplingplans.com/modern3.htm#EVALUATE
In other words, all attribute acceptance sampling plans have a zero risk of rejecting a lot if that lot contains no defectives (or defects).
The problem with setting AQL=0, alpha=0 is that since all attribute plans have oc-curves that go through this point, it is not of any value in specifying the performance of a plan.
Setting any other two points of an oc-curve will completely specify the curve and the n and Ac.
Does this answer the question, or am I misinterpreting it?
Sincerely, Stan Hilliard