Behavioural response multisession models query

[rebecca.j...@gmail.com]

Hi there,

Just wanting to confirm what type of behavioural response is implemented in oSCR.

I'm wanting to use a permanent (i.e. lifetime) change in detection probability after first capture. I.e. if captured in a session, b = 1 on any subsequent session capture too, is this how oSCR works?

Thanks,

Rebecca

[Jeffrey Royle]

hi Rebecca,

 The behavioral response is not preserved across sessions for the simple reason that individual ID is not preserved across sessions.  If you really want to preserve individual ID across sessions (which I assume are time periods like seasons or years) then you would need a custom model for that I believe, probably based on an open population Jolly-Seber type model in which the model is parameterized in terms of individual states to keep track of individual status in sequential time.  These are usually fit in Nimble or other Bayesian programs for doing MCMC.  That's a bit of a learning curve I'm afraid.

regards

andy

--
You received this message because you are subscribed to the Google Groups "oSCR" group.
To unsubscribe from this group and stop receiving emails from it, send an email to oscr_package...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/oscr_package/fcca0dcb-e939-4522-869f-2819b17080bcn%40googlegroups.com.

[rebecca.j...@gmail.com]

Thanks for confirming, Andy. That makes sense.

I'm using nimble for more complex open population models, where I have a permanent response, but here I'm trying to make some more accessible, less computationally demanding code for practitioners.