In acceptance sampling, a sample of units is randomly selected from a larger batch, lot, or process
to determine if the entire batch should be accepted or rejected. There are two main types of risks
involved:
1. Producer’s Risk:
This is the risk that a "good" batch is incorrectly rejected.
It occurs because random sampling might not always represent the batch accurately.
The acceptable quality level (AQL) is used to define what is considered a "good" quality
level, usually based on what meets the customer's standards.
For example, a producer’s risk of 5% at an AQL of 0.02 means that batches with up to 2%
defects are considered acceptable, and there is a 5% chance that a batch at this quality
level could still be wrongly rejected.
2. Consumer’s Risk:
This is the risk that a "poor" batch is mistakenly accepted.
It also occurs due to the randomness in sampling, which might miss defects in the batch.
The limiting quality level (LQL) or rejectable quality level (RQL) defines what level of
quality is considered unacceptable.
If we say that the consumer’s risk is 10% at an LQL of 0.08, this means that batches with
8% defects (poor quality) could still be accepted 10% of the time, even though this quality
level doesn’t meet standards.