• Overview of ANSI/ASQ Z1.4 (formerly MIL-STD 105) including how to define a plan based on (1) the lot size, (2) the inspection level and (3) the acceptable quality level (AQL) at which we want to accept most (nominally 95%) of all lots.
• The operating characteristic (OC) curve reflects the probability of acceptance versus the nonconforming fraction of the work.
• Any ANSI/ASQ Z1.4 plan can be converted into a zero acceptance number plan that provides equivalent protection at the rejectable quality level (RQL). While ANSI/ASQ Z1.4 plans do not have formal RQLs, we pretend for analytical purposes that this is the nonconforming fraction at which the inspection plan will reject 90% of the lots. This RQL is also used for sequential sampling plans.
   - The zero acceptance number plan has the lowest average sample number (ASN) compared to ANSI/ASQ Z1.4 single, double, multiple sampling plans, and even sequential sampling plans.
   - The operating characteristic curve shows that the zero acceptance plan has a greater chance of rejecting lots for nonconforming fractions up to the RQL, assuring the customer that it is getting better protection than the original plan. This also shows that the chance of rejecting work at the AQL is far higher than the nominal 5%, which means the producer must be very confident in the quality level to use a zero acceptance number plan.
   - Another prerequisite for a zero acceptance number plan is a decision (with the customer) as to whether ANSI/ASQ Z1.4's switching rules must be used. If there is a significant chance of rejecting good lots, the switching rules will put the producer into tightened inspection very quickly.
   -  When zero acceptance number plans cannot be used, e.g., because quality is not good enough to avoid rejection of good lots, alternatives are available.
    1. ANSI/ASQ Z1.4 double and multiple sampling plans
    2. Sequential sampling plans require somewhat less inspection than multiple sampling plans.
    3. Narrow limit gauging offers enormous reductions in inspection, but requires (1) that the quality characteristic be normally distributed, (2) the quality characteristic can be checked with go/no-go gages that can be set to specific dimensions, and (3) increases in the nonconforming fraction are due solely to changes in the process mean.

Recommended reference: Squeglia, Nicholas L. Zero Acceptance Number Sampling Plans, current edition

Attendees will receive a handout of the presentation slides and notes, along with a spreadsheet that assists in the conversion of ANSI/ASQ Z1.4 plans into zero acceptance sampling plans.