For g groups with n = 100 each, we plan to compare all pairs of population means.We want the probability to equal at least 0.80 that the entire set of confidence intervals contains the true differences. For the Bonferroni method, which t-score multiple of the standard error should we use for each interval if

(a) g = 10,

(b) g = 5? Describe how the t-score depends on g, and explain the implication regarding width of the intervals.