Skip to main content

Table 1 Moran’s I two-sided p-values for each data set using selected weight specifications defined in the previous section

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

Weight specification (Model)

Scottish lip cancer

Synthetic data 1

Synthetic data 2

Synthetic data 3

A1

0.1240

0.2619

< 0.0001

< 0.0001

A2

0.5061

0.1114

< 0.0001

< 0.0001

D2

0.2929

0.2109

< 0.0001

< 0.0001

C3

0.6504

< 0.0001

< 0.0001

< 0.0001

H5

0.6923

< 0.0001

< 0.0001

< 0.0001

  1. A p-value close to zero suggests the presence of spatial autocorrelation in the observed data
Back to article page

International Journal of Health Geographics

ISSN: 1476-072X