transients

[MSoledad]

I’ve been trying to analyze my data with a MS model with 2 unobservable states. After did GOF tests (U-CARE) I found that there’s significance in TEST3.SR (transience effect).

I’ve been constructing my models with the PIM, using the sin link for S and p, and Mlogit for Psi parameters. Now I have a TSM effect in survival rate, with a constant S for both classes, S (m2-./.) (a constant survival is an a priori hypothesis). So I’m not sure if I can continue working with the PIM (sin link and Mlogit, like before) or I have to use the DM or PIM but, in this case, with logit and Mlogit link.

I would really appreciate any help.
Thanks in advance.

MSole