Suppose that there are two statistics that will serve as an estimator for the same parameter. One of them is biased, and the other is unbiased.
a. Everything else being equal, explain why you usually would prefer an unbiased estimator to a biased estimator.
b. If a statistic is unbiased, does that ensure that it is a good estimator? Why or why not? What other considerations must be taken into account?
c. Describe a situation that might occur in which the biased statistic might be a better choice as an estimator than the unbiased statistic.