Because a sample is only part of the lot, two kinds of mistake are possible. Producer's risk is the chance that a lot of genuinely acceptable quality is rejected because the sample happened to contain a few extra defective units. Consumer's risk is the chance that a lot of poor quality is accepted because the sample happened to miss most of the defects. Every plan balances these two risks.
Larger samples reduce both risks but cost more inspection time. A lower AQL protects the consumer more but rejects more good lots. Choosing an AQL is therefore a business decision shared between factory and customer, not just a technical one.
Not every fault is equally serious, so defects are classified before counting. A critical defect could cause injury or make the product unsafe or unsaleable by law, such as a needle fragment left in a seam. A major defect would likely cause the shirt to be returned or rejected by a shopper, such as a visible hole or a mismatched sleeve length. A minor defect is a departure from the standard that a shopper would probably accept, such as a loose thread end. Each class usually carries its own AQL and its own Ac and Re.
Standard Procedure
- 1Agree written definitions and photo examples for critical, major and minor defects.
- 2Set an AQL for each class, with critical normally at zero.
- 3During inspection, record each defective unit under its most serious defect class.
- 4Judge each class against its own Ac and Re; any class reaching Re rejects the lot.
Lesson Summary
- ✓Producer's risk and consumer's risk are the two possible sampling mistakes; sample size and AQL set the balance.
- ✓Critical, major and minor defects are counted separately, and a single critical usually rejects the lot.