A Note on Maximized Posttest Contrasts

Hollingsworth recently showed a posttest contrast for ANOVA situations that, for equal N, had several favorable qualities; the contrast is maximized so that if the overall F test were significant, the contrast would also be significant. The coefficients are chosen such that , which is said to help interpret the resulting contrast. However, for unequal N, the contrast suggested by Hollingsworth fails to achieve status as a maximized contrast; thus the contrast is not insured to be significant when the overall F test is significant, requiring separate testing of the contrast.