Estimate Sample Size for very rare event


Good morning.
Suppose we have a very low proportion of a certain type of non conformity (e.g. foreign body presence). Suppose this proportion is 1/20.000. Suppose I find one NC in a lot of 1000 (proportion 1/1000) and I have to consider the possibility that in the same lot could be another NC (the presence of one foreign body could enhance the possibility to find a second NC). What is the correct approach to estimate the sample size to monitor for my lot? Minitab Sample size 1 proportion? And in that case, which values for comparison and hypothesised proportions? So, how many sample (of my lot) you would analyse to exclude the presence of a second NC with a power of at least 80%?
Top Bottom