Suppose we have a completely randomized design with 5 treatments which are randomly selected a popu-

lation of treatments. The number of replications per treatment are n1 = 6; n2 = 5; n3 = 8; n4 = 9; n5 = 7.

In computing the power of the AOV F-test, the distribution of the ratio of MSTRT to MSE is

A. central F with df=4 and 35

B. central F with df=4 and 30

C. non-central F with df=4 and 35

D. non-central F with df=4 and 30

E. none of the above