T
Hello all, I'm new to quality control (from environmental economics!) and have the following problem. We are running a project to edit existing digital maps to match the industry standard master maps. we have a target of 98% pass rate i.e 98% of maps are correct. the problem is there are lots of them - around 72 blocks of 8000 maps. we are sampling on a block by block basis.
If we look for a zero failure sample I've calculated a sample size of 149 (using
n= ln(1-(c%/(100%)/ln(1-p) where c is the confidence level (95%) and p is the probability of defective units that are to be detected (.02%). If we're on target we should have no failures in the 149? I understand that if there are failures we take a second (c2) sample to test that the number of failures is'correct' for the 98% success rate. my question, assuming this is the correct procedure, is how to calculate the size of the second sample and the acceptable number of failures in that sample.
Thanks.
If we look for a zero failure sample I've calculated a sample size of 149 (using
n= ln(1-(c%/(100%)/ln(1-p) where c is the confidence level (95%) and p is the probability of defective units that are to be detected (.02%). If we're on target we should have no failures in the 149? I understand that if there are failures we take a second (c2) sample to test that the number of failures is'correct' for the 98% success rate. my question, assuming this is the correct procedure, is how to calculate the size of the second sample and the acceptable number of failures in that sample.
Thanks.
Welcome to the cove