Dairy cows are bred by selected bulls, but not all cows become pregnant at the first service. A drug is proposed that is hoped to increase the bulls fertility. Each of seven bulls will be bred to 2 herds of 100 cows each (a total of 14 herds). For one herd (selected randomly) the bulls will be given the drug, while no drug will be given for the second herd. Assume the drug has no residual effect. The response we observe for each bull is the number of impregnated cows under drug therapy minus the number of impregnated cows without the drug. The observed differences are -1, 6, 4, 6, 2, -3, 5. Find the p-value for the randomization test of the null hypothesis that the drug has no effect versus a one-sided alternative (the drug improves fertility).