it would be relatively simple to calculate sample size based on past peel testing (variable data) with mean and stdev, against the min spec limit (1lbs) to determine sample size to meet single-sided 95% confidence 95% lower bound limit for instance as the average for most of the packaging types were high enough from the spec limit. However, for attribute data, I am having a hard time coming up with a justification that is generic enough to say since i tested 30 bubble tests (95%confidence 90% acceptable) and all passed ( - I will be able to test n=X (some lower number of samples) because ".....". Is there a reasonable way to justify less attribute samples to be tested for re-qualification?