Global model with transience in one group only.

[louise.fairless]

Dear all,

I have two groups (male and female). When I fitted the global model phi(s*t)p(s*t) in U-care, the overall GOF test was non-significant. However, the overall test for transience was significant. I would therefore be weary to continue with the presence of such transience. So, I could proceed with the more parametized transience model phi(a2*s*t)p(s*t) where there is a transience effect in both sexes.

Or, would it be more applicable to restrict transience to the female group (as this is where the transience is detected) and continue with the global model: phi[Female](a2-t/t) phi[Male](t) p(s*t) and if so, is this testable in the following way:

MALE GROUP (Test 3.sr+Test3.sm+Test2.ct+Test2.cl) + FEMALE GROUP (Test 3.sm+Test 2.ct+Test 2.Cl)

Best wishes and thank you in anticipation of your advice

Louise