What are Type I and Type II errors? How does the alpha level of a statistical test affect the risk of each type of error? How does the selection of a one-tail or two-tail test affect the risk of each type of error? How does it affect power? What would be your preference as a researcher and why?
