= Solution
The difference of the <Bernoulli distribution> variances factors as
$$
v_0-v_1=p_0(1-p_0)-p_1(1-p_1)
=(p_0-p_1)(1-p_0-p_1).
$$
The first factor is positive, whereas $p_0>1-p_1$ makes the second negative. Thus $v_0<v_1$ and $R^*=\sqrt{v_0/v_1}<1$. Therefore \b[<Neyman allocation> assigns more patients to treatment 1, the inferior treatment]. Minimizing the <variance> of a treatment comparison and maximizing the number of desirable responses are different objectives.
Back to article page
