A metallurgist has 3 potential cdf'sF1, F2, F3which she is considering using in a simulation model for eval- uating the tensile strength of an new alloy. She mea- sured the tensile strength of 45 specimens of the alloy producingY1, . . . , Y45which can be considered as iid observations on a continuous cdfF. The most appropriate procedure to use in selecting which of the 3 cdfs would provide the best representation of the process cdfFis

A. Maximum Likelihood Estimator (MLE)

B. Minimum Mean Square Error (MSE)

C. Shapiro-Wilk (SW) statistic

D. Anderson-Darling (AD) statistic

E. None of the above would be valid

ANS: