Skip to main content
. 2012 Oct 2;120(12):1727–1732. doi: 10.1289/ehp.1205006

Table 1.

Validation statistics for the KC and RS methods.

No. of validation monitors 32 13 14 4 6 5
Mean distance from monitors to estimation sites (km) 7.6 20.9 39.1 56.5 65.5 106.0
MSE
KC 1.229 1.610 1.871 0.699 1.145 2.762
RS 4.516 5.307 7.320 1.555 3.014 2.230
MSE change (%) RS to KC –72.796 –69.672 –74.438 –55.066 –62.017 19.270
MAE
KC 0.799 1.084 1.172 0.781 0.993 1.377
RS 1.551 1.883 2.264 1.019 1.626 1.279
MAE change (%) RS to KC –48.512 –42.401 –48.261 –23.384 –38.940 7.223
ME
KC 0.228 0.035 0.511 0.108 –0.586 –0.842
RS 0.202 0.402 2.264 1.019 0.148 0.088
Pearson r
KC 0.929 0.873 0.882 0.071 0.861 0.886
RS 0.733 0.534 0.879 0.644 0.447 0.908
Spearman ρ
KC 0.826 0.786 0.917 0.200 0.600 0.800
RS 0.546 0.615 0.943 0.400 0.600 0.900