5.7. For the data of Problem 5.6, do you think that the occurrences of cracks in castings would be independent, that is, if one casting has a crack, this defect is unrelated to whether the next one has a crack? Might this account for the lack of control noted? Would you still investigate points outside of limits?