I am working with a population of beetles in a relatively small area and I realized today I was using the Logit link function for the super population instead of log or identity link, which gave me very realistic estimates. When I switched to what the gentle introduction said (Super population N to log link) I am getting these crazy large estimates. Does anyone have any idea of what I am doing wrong?
Phi and recapture probability are set using logit link, Pent is using Mlogit1.