[R] ordinal response model with spatial autocorrelation

Thomas Mang Thomas.Mang at fiwi.at
Mon Aug 10 19:51:24 CEST 2009


[note: 4th posting trial - apologize if the other ones would ever show 

I have a (3-level) ordinal response data set which needs the integration 
of an spatial autocorrelation structure. What packages / functions are 
available to fit such a thing ?

The heterogeneous, cluster-alike structuring of the autocorrelation 
seems to make a mixed effects model with random factors capturing the 
different sampling regions as ideal. I failed however to find a function 
in nlme/lme4 or other packages capable of fitting ordinal data, and 
package 'DPpackage' uses an approach I am not really familiar with; the 
function DPMolmm also triggers errors, even for the provided examples.

There is no sound a-priori reason to believe that the proportional odds 
assumptions apply - they may, or may not. Anyways I don't have any 
preference to that specific kind of model, so any type of ordinal 
response model would be much appreciated !
I don't see a practical way of constructing a model out of 2 individual 
logistic regression models, but in case there is, I'd be happy to hear 
about !

BTW, I have no idea why my posts don't make it through lately. Is there 
some 'testing' newsgroup I can send dummy posts too, just to see if it 
works ?

thanks a lot and cheers,

More information about the R-help mailing list