The scientific advancement in this particular topic is not that far (since it's relatively new) but the attachment is exactly what you are looking for. Hope that helps.

Dear Avit,

Thank you very much. Yup. I totally agree with you. Kind of struggling to find the right way to do it.

Thanks again.

You might use glm with family=binomial using as covariates the proportions of the neighbours, which you could extract from you shp file (if your spatial objects are poliginals) using the function poly2nb.

Sorry, I do not read carefully the issue and do not see the word 'temporal', my suggestion was for just cross-section data.