Table 2.

Bayesian models for prevalence of Schistosoma mansoni in northwest Tanzania, based on 143 schools, 2004. DIC refers to deviance information criterion.

Variable	Coefficient, posterior median (95% Bayes. credible interval)	Odds ratio, posterior median (95% Bayes. credible interval)
a) Bayesian logistic regression model, no geostatistical component:
α (intercept)	-19.6 (-16.7 - -22.4)	-
Distance to water body <0.04 decimal degrees*	3.9 (3.5 - 4.4)	51.2 (34.4 - 78.0)
Distance to water body 0.04 - 0.1 decimal degrees*	2.3 (2.0 - 2.7)	10.5 (7.1 - 15.5)
Distance to water body 0.1 - 0.4 decimal degrees*	1.2 (0.9 - 1.6)	3.4 (2.4 - 5.0)
Annual minimum temperature °C	0.9 (0.7 - 1.1)	2.4 (2.0 - 2.9)
DIC	534
b) Bayesian geostatistical model, no covariates:
α (intercept)	-4.0 ( -5.5 - -0.9)	-
κ (smoothing parameter)	0.8 (0.5 - 1.3)	-
ϕ (decay of spatial correlation)	1.4 (0.4 - 4.1)	-
DIC	399
c) Bayesian geostatistical model with covariates:
α (intercept)	-12.3 (-18.8 - -4.5)	-
Distance to water body <0.04 decimal degrees*	4.1 (2.8 - 5.4)	59.3 (15.8 - 230.0)
Distance to water body 0.04 - 0.1 decimal degrees*	2.3 (1.2 - 3.4)	10.3 (3.3 - 31.4)
Distance to water body 0.1 - 0.4 decimal degrees*	1.1 (0.1 - 2.0)	2.9 (1.1 - 7.1)
Annual minimum temperature °C	0.4 (0.0 - 0.8)	1.6 (1.0 - 2.3)
κ (smoothing parameter)	0.8 (0.5 - 1.3)	-
ϕ (decay of spatial correlation)	2.8 (1.0 - 5.7)	-
DIC	397

^*reference category: Distance to perennial water body (>0.4 decimal degrees)