When we want to compare one main effect to another main effect, we can no longer use a t-test. What is true?
1) A t-test can still be used to compare main effects
2) A different statistical test should be used to compare main effects
3) Main effects cannot be compared
4) Comparing main effects is not necessary