I am rather new to MARK, and I was wondering if for Known-fate models it is appropriate to fix parameters to 1 in time-dependent models for time intervals where no mortalities are known to have occurred?
It is my understanding from section 16.4 of the MARK Gentle Introduction book and the worked black duck example that the number of model parameters estimated when using the design matrix and logit function should be readjusted to include those parameters where survival was estimated at 1. However, some journal articles that have used Known-fate models have described in their methods that they fixed parameters to 1 when no mortalities were known to have occurred in an interval (e.g. Jones et al. 2012, "Range‐wide analysis of eastern massasauga survivorship"). Which method is the "correct" way to go about accounting for such intervals in time-dependent KF models in MARK?
Any advice at all would be much appreciated, and thank you for your time!