Mean separation letter from test slices

[#1]

Hello!

How do I get the mean separation letter in for each level of a factor when interaction is significant? For example, if I have A*B interaction and I want to get the mean separation of B at each level of A, how do I get the LSD letters in JMP?

I can do test slices and that will use the DF of the full analysis, but this does not shows which mean of B differs.

If I do multiple comparison:

How can I get the mean separation letters from here? Does this modify the P-Value to interpret the final conclusion (Slicing for every level of A = 0.05/2, and then getting a test for every level of B in each level of A =6?

I hope I express clear my questions, and I will appreciate any help or comments on how to interpret this analysis?

Thanks in advance,