sorry if the question is obvious, but nooby alert!
I am running a full factorial model with Factor A, categorical 2 levels, Factor B, categorial 7 levels, and Factor C (participants) defined as a random factor.
I use the standard least square personality, and I have JMP. 15.
I have found a signiticant interaction between the factors A and B, and now I want to do 7 post-hoc t-tests between all the individual levels of B: i.e.
A1-B1 vs A2-B1;
A1-B2 vs A2-B2;
...
A1-B7 vs A2-B7;
if I used the "LSmean Students t" tools on the interaction, it will compute all possible pairs (including A1-B1 vs A2-B2, ...). This will have a consequence on my statistical power. So how can I restrict the t-test to just the one that I need?
Many thanks
Jennifer.
