What is it about?
We present a new method based on distances, which allows the modelling of continuous and non-continuous random variables through distance-based spatial generalised linear mixed models. The parameters are estimated using Markov chain Monte Carlo maximum likelihood. The method is illustrated through the analysis of the variation in the prevalence of Loa loa among a sample of village residents in Cameroon.
Featured Image
Why is it important?
We describe a refined spatial generalised linear mixed model incorporating general measures of distance/dissimilarity that can be applied to explanatory variables: numerical,categorical or a mixture of them. Methods of inference for such models (including simulated-based maximum likelihood estimation for model fitting and a discussion of methods for model comparison and testing goodness-of-fit) are provided.
Perspectives
The resulting models are termed generalised linear mixed models (GLMMs), providing a convenient and flexible way to model multivariate non-normal data. In particular, GLMMs constitute a unified framework for modelling geostatistical non-normal data, using mixed terms to model the underlying spatial process. We present a new method based on distances for modelling spatial GLMMs. This model is useful when the explanatory variables are not necessarily continuous or they are a mixture of continuous and categorical variables. Our distance-based spatial generalised linear mixed model method is used to predict the trend and to estimate the covariance structure when the explanatory variables are continuous, categorical, binary or a mixture of them.
Professor Oscar O. Melo
Universidad Nacional de Colombia
Read the Original
This page is a summary of: Spatial generalised linear mixed models based on distances, Statistical Methods in Medical Research, December 2013, SAGE Publications,
DOI: 10.1177/0962280213515792.
You can read the full text:
Contributors
The following have contributed to this page
