If you have a pool of 20 subjects (and plan to use all subjects for your study), how many subjects per group would you have for the following experimental (balanced) designs?

(i) S x A x B with 2 levels in A and 2 levels in B:

(ii) S(A x B) with 2 levels in A and 2 levels in B:

(iii) S(A) with 4 levels in A:

(iv) S x A with 5 levels in A: