ABSTRACT
Let X1, . . . , Xn be a sample of independent observations from a continuous population. Let M denote the median of the population. We want to test
H0 :M =M0 vs. Ha :M 6=M0.
Let K denote the number of plus signs of the differences X1−M0, . . . , Xn−M0. That is, K is the number of observations greater than M0. Then, under the null hypothesis, K follows a binomial distribution with number of trials n and success probability
P (Xi > M0) = 0.5.
