can you get residuals from occupancy models?

questions concerning analysis/theory using program PRESENCE

can you get residuals from occupancy models?

Postby Lea » Thu Dec 03, 2015 6:06 pm

Hi there,
I am concerned that spatial auto-correlation may be a problem in my study area but not such a problem that I think I need to resort to using the spatial dependence models. Is there anyway to get residuals for multi-season models so that I can assess this?
Thanks!
Lea
Lea
 
Posts: 33
Joined: Thu Oct 24, 2013 1:09 pm

Re: can you get residuals from occupancy models?

Postby jCeradini » Sat Dec 05, 2015 3:16 pm

Hi Lea,

I'm not sure about the residuals question - hopefully someone more experienced will address that question.
Alternatively, could you do a more informal test for something like a "block effect" by having models with and without that term in your candidate set? So, if there are groups of sites or plots that are closer together in space or were sampled at a similar time, you could group those sites into blocks as a new predictor, and see if that term improves model fit and/or is significant.

Joe
jCeradini
 
Posts: 72
Joined: Mon Oct 13, 2014 3:53 pm

Re: can you get residuals from occupancy models?

Postby jhines » Sat Dec 05, 2015 8:38 pm

By 'residuals', I assume you mean cell probabilities times the number of sites. The problem with multi-season occupancy models is that typically you'll have many surveys and the cell probabilities will be very small for many detection-histories. So, you'll be comparing a single detection-history with a residual which is tiny. That makes it tough to make an assessment about 'fit'.

My suggestion would be to run the multi-season correlated detections model and the standard multi-season model and let AIC tell you if the correlated detections model is needed.

Jim
jhines
 
Posts: 599
Joined: Fri May 16, 2003 9:24 am
Location: Laurel, MD, USA

Re: can you get residuals from occupancy models?

Postby Lea » Tue Jan 19, 2016 5:23 pm

thanks for your responses Joe and Jim. Are there any worked examples or exercises for the correlated detection models? I was having trouble trying to figure out how to input the data and set up the DM which is one reason I steered clear of it.
Thanks,
lea
Lea
 
Posts: 33
Joined: Thu Oct 24, 2013 1:09 pm


Return to analysis help

Who is online

Users browsing this forum: No registered users and 9 guests