1. Why do we generally need to correct for the probability of making Type I errors for post hoc, but not a priori comparisons?
2. What is the average cell standard deviation in Factorial anova?
3. how to calculate power in one way anova?
4. What is the advantage of doing the SS contrast above by hand, rather than computing these simple effects analyses in SPSS?
here is the answer(how to understand?):
It is more powerful because we get to use the larger df from the omnibus ANOVA which lead to lower critical values that need to be exceeded. Because you use the MSE from the omnibus ANOVA, your estimate of error variability is more accurate estimate of the population variance because it is averaged over more groups.
