Dear Gonzalo, Dr Kitron et al.,
I have a doubt about some of the coefficients in Table 3. Specifically, the coefficients for Density of houses and NDVI look extremely large. If these are the usual beta coefficients from logistic regression, then exp(beta) estimates the odds ratio. But exp(40.4)=351,153,554,528,347,000 (for each extra house/sqr km!!), and exp(59.76)=89,833,483,037,290,800,000,000,000 (!!!)... The other coefficients seem to make sense, e.g., exp(-1.48)=0.228, or exp(0.91)=2.48). Maybe I missed something, but it looks like there might be something wrong with the model - perhaps it's worth checking.
Un saludo,
Fernando Abad-Franch
