"Suppose that we wish to test the null hypothesis, H0, that the proportion of ledger sheets with errors, p is equal to 0.05 against the alternative Ha, that the proportion is larger than 0.05 by using the following scheme. Two ledger sheets are selected at random. If both are error free, we reject H0. If one or more contains an error, we look at a third sheet. If the third sheet is error free, we reject H0. In all other cases, we accept H0.
a) What is the value of A [Type I Error]
b) Calculate the type II error B as a function of p.
I am completely lost. The scheme sounds like utter nonsense. Even if only the third sheet has an error, that is a 0.33 proportion, which is much greater than 0.05. I feel like this sample size is way too small to draw any conclusions about A and B... Help please!