of the average pulse rates at KMN and KS: Thedata are compatible with a geographic trend.(b) 9*8/2 = 36 comparisons.(c) The honest significant difference from Tukey’s methodisdHSD= 5.21*√567*1/15 = 1.013. (usingk= 9 groups and df Error≈120 in the table).The observed difference between KW1 and KW2is 0.18, which is smaller thandHSD.Therefore,no:KW1 and KW2 would not be declared tohave significantly different means.(d) =α*=.01/36 thenα*/2∼.00014 on each side,for 2-sided pair-wise comparisons.(e) Fisher’s least significant difference isdLSD= 2.62*√567*2/15 =.72, which is smaller thandHSD=1.013.So any pair of locations whose averagedifference is between.72 and 1.01 would be de-clared significant based on Fisher’s LSD, but non-significant based on Tukey’s HSD. KMN - KW1=.92 is such a pair.(f) Bonferroni is the most appropriate in this case,because it has the lowest type I error rate. Therewill be few significant comparisons with this method(which is okay with the follow-up study), but ofthese, we should be very confident that they arereal (which is highly desirable for the follow-upstudy).1
This is the end of the preview.
access the rest of the document.