4.15. For the artificial data sets generated for Problem 1.6, compute and compare estimates of the mean and variance of Y, from the following methods:

(a) Complete-case analysis;

(b) Buck's method, imputing the conditional mean of Y given Y from the linear regression based on complete cases;

(c) Stochastic regression imputation based on the normal model, where a random normal deviate N(0.5) is added to each of the conditional means from (b);

(d) Hot-deck imputation, with adjustment cells computed by categorizing the complete cases into quartiles based on the distribution of Y. Suggest a situation where

(d) might be a superior method to (c).