Key points are not available for this paper at this time.
In many industrial environments it is reasonable to assume that batches of items produced sequentially will be positively correlated. Taking advantage of this assumption it is shown how the sentencing of a batch as good or bad can depend upon the number of defectives in a random sample of fixed size drawn from the particular batch to be sentenced and upon the number of defectives found in batches before and after.
D. R. Cox (Mon,) studied this question.