A researcher is planning to conduct an experiment to compare two treatments in which matched pairs of subjects will be given the treatments and a sign test will be used, with a nondirectional alternative, to analyze the difference in responses. Suppose the researcher believes that one treatment will always do better than the other. How many pairs does he need to have in the experiment if he wants to be able to reject H0 α = 0.05? when If one treatment "wins" in every pair, what will be the P-value from the resulting test?