Suppose you performed an ANOVA with a = 5 treatments and n = 4 observations per treatment. If the MSE = 16 and lpha = 0.01 (significance level), what would the minimum difference have to be between any two sample means for you to conclude that the population means were significantly different if you performed

(a) The LSD comparison procedure?

(b) The Bonferroni comparison procedure?

(c) Tukey's multiple comparison procedure?

(d) Scheffe's procedure?

(e) Explain the relationship between power and the minimum difference. Also state which of the above four is the most powerful and least powerful comparison procedure.