Geoadditive models



Matt Wand

Harvard School of Public Health


A study into geographical variability of reproductive health outcomes (e.g. birthweight) in Upper Cape Cod, Massachusetts, benefits from geostatistical mapping or kriging. However, also observed are a number of continuous covariates (e.g. maternal age) that exhibit pronounced non-linear relationships with the response variable. To properly account for such effects we merge kriging with additive models to obtain what we call geoadditive models. The mergence becomes effortless by expressing both as linear mixed models. The resulting mixed model representation for the geoadditive model allows for fitting and diagnosis using standard methodology and software.

Back to statistics seminar home page