(3) Suppose you have an independent sample of two observations, denoted 1 and y, from a population of interest. Further, suppose that E(y) = and Var(= 0%, i = 1,2 Consider the following estimator of : i = c + dys. С for some given constants c and d that you are able to choose. Think about this question as deciding how to weight, the observations y and y2 (by choosing c and d) when estimating (3a) Under what condition will ſo be an unbiased estimator of ye? (Your answer will state a restiction on the constants c and d in order for the estimator to be unbiased). 3 (31) Given your answer in (3a), solve for din terms of cand substitute that result back into the expression for janbove. Note that the resulting estimator, now a function of c only, is unbiased Once you have made this substitution, what is the variance of je in terms of o' and d? (30) What is the value of that minimize the variance expression in (3b)? Can you provide any intuition for this result? (34) Re-derive the variance in part , but this time suppose that Var() = ? and Var) = 207 If the variances are unequal in this way, what is the value of that minimize the variance expression? Comment on any intuition behind your result