Suppose you wanted to construct 10 confidence intervals with a family wise error rate of 5% (or confidence intervals with a 95% simultaneous confidence level). Suppose furthermore that the confidence intervals were independent (not the case when testing all pairwise comparisons).
a) Design an adjustment procedure for this situation and explain why it works.
b) Now generalize to the situation where you have n tests and want a family wise error rate of α.
c) How does this method compare with the Bon ferroni correction?