Page 1 of 1

Multi-method model: Calculating residuals for spatial auto?

PostPosted: Mon Nov 27, 2017 9:46 am
by BrynEvans
Hello all, I have another question for multi-method models!

I would like to assess spatial auto-correlation, and so I am in search of the best way to calculate residuals. I have "stations" with different methods, but some of these whole stations are much closer together than others (though the methods within each are consistent). However, they don't fall into a strictly spatial-dependence category, nor am I sure there is an appropriate hierarchical approach I could use. Any advice would be much appreciate.

My very first approach was to use the multi-method models in PRESENCE, determine meaningful covariates for p and psi, and then run a simple single-season model with only data from the best METHOD, and look at those residuals. If there is something I can do directly at the multi-method stage though I would value any insight!

Thanks for your time,
Bryn