A pharmaceutical firm produces capsules that are supposed to contain 50 mg of a certain drug. The precise weights vary from capsule to capsule but can be assumed to have a normal distribution.
(a) Suppose that one is interested in constructing a 95% confidence interval that would estimate the mean weight of capsules to within 2 mg, and you suspect that a is no larger than 3 mg. How large a sample size is necessary?
(b) A random sample of 16 capsules results in the statistics x = 47 mg and s = 4 mg. Construct a 90% confidence interval for the mean weight of the capsules produced by this process.
(c) How large should a second sample be in order to estimate the mean weight of capsules to within 1.5 mg with 95% confidence?
(d) If the second sample resulted in x = 44 and s = 4.5 mg, what would you obtain for the 95% confidence interval?