GEE with temporal and spatial correlation R

#1
Hello,

I am working with a dataset with temporal correlation (multiple rounds; ~1050) and spatial clusters (correlation by place ~20).

I am planning to fit a GEE model using R "geepack", but I'm uncertain if the "id" parameter should be set to the spatial or temporal cluster. Or should I use a MLM because I have two levels of correlation.

Thanks!
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
Cool dataset. I am not seasoned in this area. I am guessing you will have put one of them as a term since high complexity can get out of control. I know in spatio regression you can get a spatial term. Then perhaps put that term in your MLM. Remember i am not experienced with dual types of data.