When computing individual confidence intervals using the t statistic for all possible pairwise comparison of means,the experimentwise error rate will be:
A)Equal to $\alpha$
B)Less than $\alpha$
C)Greater than $\alpha$

Question

Quizplus · Accepted Answer

When computing individual confidence intervals for all pairwise comparisons of means, the experimentwise error rate increases due to inflation of Type I error. Therefore, the true alpha-level for the entire family of tests will be greater than the nominal level (Sα). The formula for experimentwise error rate is FWE = 1 - (1-α)^nC2, where n is the number of samples/groups and C2 is the number of pairwise comparisons. As the number of pairwise comparisons increases, the experimentwise error rate (FWE) also increases. Therefore, the best choice is C, which states that the FWE will be greater than Sα.

When Computing Individual Confidence Intervals Using the T Statistic for All