(Solved): Please upload your answers to the following questions (show all your work): A graduate student used ...
Please upload your answers to the following questions (show all your work): A graduate student used a four-group between-subject design for her thesis. She had 31 subjects per group. Her sample means were group
1=22, group
2=18, group
3=40, and group
4=44. The value of MS Within was 1600. (6 pts) (a) Should she reject an omnibus null hypothesis? (2 pts) (b) She decides to test all pairwise differences, assuming homogeneity of variance and using an appropriate method for controlling experiment-wise error rate. Does she obtain any statistically significant differences? (6 pts) (c) After looking at the data, her adviser asks her to compare the average of groups 1 and 2 with the average of groups 3 and 4 , again controlling for experiment-wise error rate. She argues that performing this test is fruitless because tests of complex comparisons are more conservative than tests of pairwise comparisons. Is this complex comparison statistically significant? Is the student's reasoning correct? Show your work.
