Skip to main content

Table 2 Global spatial autocorrelation of risk factors and their correlation with malaria incidence

From: Geo-visual integration of health outcomes and risk factors using excess risk and conditioned choropleth maps: a case study of malaria incidence and sociodemographic determinants in Ghana

Risk factors Global spatial autocorrelation #Pearson's correlation with outcome rate Bivariate spatial correlation with outcome rate
Univariate Moran's I Pseudo p-value z-value Pearson's r p-value Bivariate Moran's I Pseudo p-value z-value
Basic education 0.138*** 0.019 3.193 -0.226* 0.003 -0.062*** 0.042 -1.493
Illiteracy 0.013 0.056 1.650 -0.103 0.180 -0.0004 0.465 -0.069
Religion
 Christian -0.009 0.275 -0.072 -0.057 0.457 -0.019 0.329 -0.491
 Muslim 0.277*** 0.002 5.186 -0.137 0.075 -0.039 0.140 -0.971
 Traditional 0.657*** 0.001 10.855 0.014 0.856 0.019 0.324 0.414
 None/Other 0.176*** 0.010 3.671 -0.194* 0.011 -0.064*** 0.046 -1.505
Urban lev. 0.117*** 0.022 2.860 -0.174* 0.023 -0.053 0.070 -1.287
Insanitary lev. 0.018 0.080 0.881 -0.002 0.983 0.008 0.374 0.226
Intermigration -0.001 0.125 0.828 0.184* 0.016 0.089*** 0.038 1.969
Intramigration 0.102*** 0.034 2.100 0.144 0.060 0.111*** 0.020 2.402
Traditional housing unit 0.030 0.129 0.849 -0.122 0.112 -0.021 0.314 -0.523
Household overcrowding index 0.076*** 0.031 2.112 -0.148 0.054 -0.034 0.190 -0.836
Pop. density 0.018 0.080 0.881 -0.002 0.984 0.008 0.374 0.227
Dependency ratio 0.550*** 0.001 9.051 -0.020 0.795 0.070 0.050 1.627
Employment to-population ratio 0.185*** 0.012 4.777 -0.189* 0.014 -0.061*** 0.042 -1.503
Agric household 0.122*** 0.021 2.8055 0.178* 0.020 0.115*** 0.015 2.536
  1. * significance level at p< 0.05 (2-tailed); ** significance level at p< 0.01 (2-tailed)
  2. ***spatial significance level at pseudo p < 0.05 for conditional 999 permutations.
  3. # Spatial correlation is most appropriate but the spatial correlation was much smaller than the non-spatial correlation and the departure from independence is consistent but weak. The assumption of spatial independence may have affected the non-spatial correlation result.