Skip to main content
. 2007 Sep 20;6:39. doi: 10.1186/1476-072X-6-39

Table 2.

Estimation of the ecological link β when E¯h MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWGfbqrgaqeamaaBaaaleaacqWGObaAaeqaaaaa@2F5C@ = 23.35 and ρxx = 0.0, 0.4 (400 replications)

ρxx(1) β Model β¯ MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaiiGacuWFYoGygaqeaaaa@2E6C@sim(2) σ¯βsim MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaiiGacuWFdpWCgaqeamaaBaaaleaacqWFYoGydaWgaaadbaGaem4CamNaemyAaKMaemyBa0gabeaaaSqabaaaaa@34BB@ MB(4) sd(MB) sd(βsim)(5) RMSE(6) π(β)(7) 1 - π(0)(8)
0.0 0.00 M0 0.000 1.63 1.73 1.73 0.953 0.047
BYM 0.000 1.76 1.74 1.73 0.958 0.042
0.12 M0 0.119 1.61 -0.44 0.69 1.67 1.67 0.930 1.000
BYM 0.120 1.72 -0.31 0.70 1.68 1.68 0.955 1.000
0.21 M0 0.210 1.60 0.49 0.39 1.64 1.64 0.948 1.000
BYM 0.210 1.72 0.58 0.39 1.66 1.66 0.960 1.000
0.33 M0 0.329 1.57 -0.28 0.23 1.58 1.58 0.953 1.000
BYM 0.329 1.70 -0.24 0.24 1.58 1.58 0.968 1.000

0.4 0.00 M0 -0.001 1.72 1.76 1.76 0.955 0.045*
BYM -0.001 1.90 1.78 1.78 0.975 0.025
0.12 M0 0.121 1.73 0.79 0.73 1.75 1.75 0.945 1.000
BYM 0.121 1.91 0.78 0.74 1.78 1.78 0.968 1.000
0.21 M0 0.211 1.72 0.41 0.39 1.66 1.66 0.958 1.000
BYM 0.211 1.90 0.50 0.39 1.67 1.67 0.973 1.000
0.33 M0 0.332 1.70 0.71 0.25 1.68 1.69 0.945 1.000
BYM 0.332 1.89 0.70 0.25 1.70 1.71 0.973 1.000

(1) ρxx : autocorrelation at 100 km

(2) β¯ MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaiiGacuWFYoGygaqeaaaa@2E6C@sim : 100 * mean of posterior means

(3) σ¯βsim MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaiiGacuWFdpWCgaqeamaaBaaaleaacqWFYoGydaWgaaadbaGaem4CamNaemyAaKMaemyBa0gabeaaaSqabaaaaa@34BB@ : 100 * mean of posterior standard deviations

(4) MB : 100 * Mean Bias

(5) sd(βsim) : 100 * standard deviation of posterior means

(6) RMSE : 100 * Root Mean Square Error

(7) π(β) : Coverage proportion β ∈ CI(β)

(8) 1 - π(0) : Non-Coverage proportion 0 ∉ CI(β)

(*) : Indicator of p-value < 5% from McNemar's test (1)(2)(3) for comparing proportions 1 - π(0) under M0 and BYM models.