You can find information on AOQ (or Average Outgoing Quality Level) in "Statistical Quality Control" by Grant/Leavenworth. I always use this tool in evaluating the selection of an OC curve on which to base a sampling plan.
If you are not familiar with the topic, I highly recomment reading both the sections on AOQL but also include those regaring OC (Operating Characteristic) curves. You will find, for instance, that FIXED SAMPLE SIZE TENDS TWOARD CONSTANT QUALITY PROTECTION. This is an extremely difficult concept to explain to advocates of variable sample size schemes (ie: MIL-STD 105 or using fixed percentage of the lot size for the sample size). The concepts are counter-intuitive but statistically valid. I always used a fixed sample size (despite the lot size) and determine what that is by using the worst case AOQL.
