E. canis infection is a tick-borne disease of dogs that is sometimes con-tracted by humans. Among infected humans, the distribution of white blood cell counts has an unknown mean m and a standard deviation s. In the general population the mean white blood count is 7250 per mm3.
It is believed that persons infected with E. canis must on average have a lower white blood cell count. What is the null hypothesis for the test? Is this a one- or two-sided alternative?