Skip to main content

Table 3 Parameter estimation results for observed 2004 WNV death percentages.

From: A comparison of six analytical disease mapping techniques as applied to West Nile Virus in the coterminous United States

  Gaussian SAR Gaussian spatial filter binomial spatial filter
Statistic estimate se estimate se estimate se
Spatial autocorrelation 0.01 0.101 1 (spatial filter MC = 0.030) 0.135 6.07 4.701
intercept -3.30 0.172 -3.31 0.124 -3.48 0.182
I0 2.63 0.491 2.62 0.323 -20.01 41223
approximate dfs 45 39 45
Residual P(S-W) 0.249 0.256 < 0.0001
Residual MC -0.0224 -0.0753 (zMC = -0.3) -0.073
Residual GR 1.1361 1.0788 1.088
deviance 0.9001 0.4686 1.0346
pseudo-R2 0.494 0.754 0.07
predicted-observed regression slope 1.31 0.72 0.87
HGLM
  proper CAR spatial filter
Statistic estimate se estimate se   
spatial autocorrelation -0.08 0.148 0.999 (spatial filter MC = 0.762) 0.226
intercept -3.28 0.122 -2.99 0.122
I0 -81.63 62 -80.94 60
approximate dfs 45 44
residual P(S-W) <0.0001 0.192
residual MC 0.0596 0.0130
residual GR 0.9417 0.9270
deviance 0.9715 0.4992
pseudo-R2 0.328 0.411
predicted-observed regression slope 30.60c 0.99