Dear list users,
To begin, I apologise for any naivety displayed in this post! I'm still trying
to come to terms with some of the ideas in spatial regression. I hope someone
can provide me with some guidance on a problem I have.
I am analysing data on individuals nested within zones so automatically used
multilevel modelling to reflect this hierarchy. I have both individual level
and zone level variables to use to determine an individual level response. All
zones are found within a city and some of the zones for which I have individual
level information are neighbours. I have a matrix of 0s and 1s to reflect this
information so I was wondering if it possible to fit some sort of spatial
multilevel model to this data in R? I am not sure that this is necessary as it
does not appear that the random slope effect for zones is significant in the
model. However, I retained this effect as some of the zone level variables are
statistically significantly associated with the response.
Is there some sort of a test to assess whether or not there is residual spatial
autocorrelation which has not been taken into account by my multilevel model? I
tried extracting the zone level residual information and using a moran test.
However, I have a feeling that this may not be the correct approach to take.
Anyone have any suggestions?
Cheers,
Karen
[[alternative HTML version deleted]]
_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
