Suppose that a random sample of 20 power window mechanisms is taken from a lot supplied to an auto manufacturer. Each sampled mechanisms is tested by putting it through continuous up-down cycles until it fails. Suppose that in the entire lot, the mean time to failure is 4200 cycles and that the standard deviation 3400. The mean failure time for the sample is recorded.
- Would it be reasonable to assume that the distribution of individual failure times should be roughly normal?
- Would it be reasonable to assume that the distribution of sample means should be roughly normal?