if we wanted to compare the averages of each set of scores, what type of statistical test would you run and why?