I'm doing a MANOVA two factors on SAS and I have a few doubts. Does anybody can help me?
1. The software takes all possible 2-2 comparisons, but I want just some of them. I'm trying to get a smaller confidence interval since the significance level (alpha) depends on the number of comparisons! How can I ask for the comparisons I'm interested?
2. Is the confidence level to Bonferroni Confidence Intervals always 95%? Does the alpha depend or not on the number of comparisons?
3. When the software indicates that the comparison is significant at the 0.05 level, is it about the t-value? How can I ask for t and p-values?
4. The note "This test controls the Type I experimentwise error rate" is related to the t-test, isn't?
Thank you! Renata Matos