Hi
I'm wanting to reconcile results from my own code with MARK results from models such as "Robust Design Pradel Recruitment Huggins' Closed Populations". I get agreement when all primary intervals are 1.0 and not otherwise.
I'm fairly clear after searching the MARK book that only Phi is scaled by interval duration in MARK, and I can see there will generally be problems with multi-state and robust design models. However, the adjustment seems fairly straightforward when the f parameterisation is used for recruitment (as in Link & Barker):
f' = (Phi + f)^t - Phi^t
to ensure lambda' = Phi' + f' over an interval of duration t. Does this sound right? Has this been considered and rejected...?
Murray