For your convenience, the data from Exercise 4 from the previous chapter are reprinted in the following table:

a. Using the harmonic mean of all the sample sizes as your value for n, calculate HSD for these data, and determine which pairs of groups differ significantly.

b. Recalculate HSD according to the rules of the Fisher-Hayter test. Would the use of the F-H test be justified in this case? Assuming the F-H test is justified, what conclusions can be drawn from this test? Use the values for HSD in this part and part (a) to compare the power of Tukey's test with the modified LSD test.

Transcribed Image Text:

Group 3 4 5 23 29 30 7.2 15 34 26 7.0 14 6.5 5.8 12 6.0 15 N 12