Skip to main content
Fig. 8 | International Journal of Health Geographics

Fig. 8

From: Spatial smoothing in Bayesian models: a comparison of weights matrix specifications and their impact on inference

Fig. 8

Spatial representation of the log-relative risk surface, and a breakdown of its main components: the structured and unstructured spatial random effects and the area-specific covariate effect, i.e. \(\beta x_{i}\). The values are the posterior means. The two areas shaded grey contain values larger than 5 standard deviations from the mean value. The shadings are consistent across all the plots within a given data set, and the legend shows the range of values for the specific parameter

Back to article page