Visitation Effect Issue- Nest Survival

[dnonne]

Greetings all,

Quick question...

I am looking to model visitation/flushing effects in a nest survival framework, using the basic idea that Rotella et al. explained. I have the input file set up for a 42 day nesting season, each hen starts at day 0, in order to model nest age in a more simplistic manner. 42 visitation and 42 flushing covariates, ect.

However, my concern is that I am modelling the flushing effect on the wrong day. I might be way off base here, but if the flushing effect takes place on the day that the nest is flushed, which would be the date of last activity on nests that fail due to the flushing, the model wouldn't allow for the nest to fail on that date, since it is attributing the failure to the midpoint. And this would result in the flushing covariate not to be well supported in model selection.

Would offsetting the flushing covariate effect to the day after the flush took place be more appropriate, or am I completely wrong here?

Thanks in advance

[Rotella]

The mid-point assumption is not used in the nest-survival models in MARK. So, I think your concern isn't necessary. I think that some of the supporting literature on the modeling approach in MARK could help here, e.g., Dinsmore et al.