1. a) What is a big disadvantage of cluster sampling? b) 0600930 = @ j) k) 1) m) Discuss the relative merits of the ordinary sample mean , ratio, regression and difference estimators. I.e. when is each most efficient (precise), under what circumstances should they not be used and why, etc. Under what conditions does stratified random sampling have its greatest precision? Under what circumstances does cluster sampling have its greatest precision? When is Neyman allocation optimal? When is proportional allocation optimal? Why might one use proportional allocation even if it is not optimal? Why might one use Neyman allocation even if it is not known to be optimal? What is meant by an unbiased estimator? Which of all the estimators that we have studied are unbiased? Under what circumstances are the biases in the ratio & regression estimators small? Explain the difference between accuracy and precision. Why it may sometimes be more efficient to use a biased estimator? For cach of the following situations would you use: y, a ratio estimator, a regression estimator, a difference estimator to estimate the population mean?