Post-hoc comparisons 6. What is the critical value for the Tukey-Kramer q statis
ID: 3311637 • Letter: P
Question
Post-hoc comparisons
6. What is the critical value for the Tukey-Kramer q statistic for a post-hoc comparison of mean values in this study? Explicitly state whether you used the tables in the lab manual or in the textbook. Again, include all relevant information (including name of statistic, degrees of freedom, alpha level, and number of tails)
7. Comparison of which mean values provides the most direct assessment of whether dogs affect the number of birds in the area? Why?
8. Calculate the observed q statistic for this comparison. Show all your work.
9. What do you conclude?
Mean # birds per site | SD Disturbance type None (Control) Humans only Humans with dogs 10.1 2.74 28 9.7 3.25 30 7.3 3.52 30Explanation / Answer
here level of significance alpha=0.05 is not given let alpha=0.05
(6)there are 3 group or treatment and error df=85 so q-statisitc is q(0.05,3,85)=3.71 (two decimal approximation) using q-table on internet. this is two tailed
(7)critical range=wij=q(alpha,k,n-k)*sqrt((MSE/2)*(1/ni+1/nj))=3.71*sqrt((10.216/2)*(1/ni + 1/nj))
(7) comparison of 2 and 3, are significant as difference=2.4 is more than wij=2.16 , so dogs affect the number of birds in the area
(8) observed q-statistic for calculation of difference wij for this comparison is 3.71
(9) comparison between 1 and 2 are not significant, and 2&3 and 1&3 are significant.
pariwise q(0.05,2,85) sqrt(mse/r)*(1/ni+1/nj)) wij |Xi-Xj| 1&2 3.71 0.593881502 2.2033 0.4 1&3 3.71 0.593881502 2.2033 2.8 2&3 3.71 0.58355234 2.164979 2.4