Note the following:
• Values are on the -log10 scale, so 1.3 corresponds to a p-value cutoff of 0.05, 2 to 0.01, 3 to 0.001, and so on.
• If you do not specify LSMeans differences or ESTIMATE statements (where applicable), the selection is made across Type 3 tests of the fixed effects.
•
The -log10(p-value) Cutoff field becomes active.