atements-are-correct-about-significance-and-how-many-pairwise-nulls-are-rejected

Question

A study with k = 4 groups and error df = 20 uses alpha = 0.05. You have computed q statistics: 3.10, 2.85, 4.20, 1.75, 3.95, 2.10 (six pairwise comparisons). If q_critical from the q table is 3.77 for these parameters, which statements are correct about significance and how many pairwise nulls are rejected?

Accepted Answer

Reject the null for comparisons with q = 4.20 and q = 3.95 (two rejections); the others are not significant because they are below 3.77.

Answer

Fail to reject all nulls because the familywise critical threshold must be doubled when there are six comparisons, so none reach the adjusted threshold.

Answer

Reject the null for q = 3.10, 3.95 and 4.20 (three rejections) because any q above 3.00 is considered significant for k = 4 regardless of df.

Answer

Reject the nulls for all q greater than the mean of the q statistics because the studentized range procedure compares each q to the sample mean rather than a tabled critical value.