Did anyone deal with a need to include
spatial correlation into logistic models?
I am running logistic regressions to
determine significant factors affecting distribution crop yield categories within
fields.
The yield observations tend to spatial
correlate within fields that validates the assumption that the response
variables are independent.
There is an Option in Spatial
Regression to consider spatial or covariance structure by building a neighbor
grids for ordinary linear regression.
Is it possible to do the same with
Logistic Regression?
Thanks,
Peter