We are replicating data with a new process. We don't have time to compare all the data afterwards, but we could compare a random sample of 20%. We found 1 error. The boss says there could have been hundreds of errors, but this does not seem likely to me. Based on the sampling we do if I get 1 error in 20%, how can I calculate the chances any number of X errors in the remaining 80%
Comments
Leave a comment