1. Suppose that we have a binomial response with expected failure probabilities z'o : 0.2 and an : 0.1 for treatments 0 (control) and 1 (experimental) respectively. Calculate the required sample sizes to achieve 90% power (type II error rate of 10%) at level as = .01 assuming (a) (10 pts) Equal allocation to groups 0 and 1 (b) (10 pts) 2:1 randomization to groups 0, 1 respectively ) l (c (d (10 pts) Comment on any differences or similarities among the required sample sizes? (5 pts) 2:1 randomization to groups 1, 0 respectively