

Choose your goal for multiple comparisons. On the next tab, you can choose exactly which test to use.
In many cases, the answers you really want are determined by multiple comparisons and not by the overall ANOVA analysis. Choose the comparisons you want to make on this tab, and then set the details on the next (Options) tab.
Be thoughtful about which comparisons you need to make as part of your experimental design. If you make too many comparisons, the tests become stricter and it is harder to find anything.
If you have 2x2x2 ANOVA, then the number of comparisons are:
•Compare each cell mean with every other cell mean: 28 comparisons
•Compare the control mean with the other cell means: 7 comparisons
•Compare means that differ by only one factor: 12 comparisons
•Compare each mean in row 1 with the cell mean just below it: 4 comparisons
If you have 2 x 2 x k ANOVA, then the number of comparisons are:
•Compare each cell mean with every other cell mean: 8k22k comparisons
•Compare the control mean with the other cell means: 2x2xk1 comparisons