Referring to example 8 in this section, compare ethnic groups A and B, ignoring type of medication. Assume this is the only hypothesis to be tested and that the goal is to have a Type I error probability of .05.