Skip to main content

Advertisement

Table 5 A multivariate hurdle GLM showing important factors explaining variations in epg between Kenyan elephant populations

From: Patterns of helminth infection in Kenyan elephant populations

CovariateEstimateSEZ-valueP-value
Count model coefficients (truncated negbin with log link)
 Intercept3.090.923.350.001
 3-month mean NDVI24.1810.042.410.016
 Sub-adults and Juveniles vs Adults0.110.180.630.528
 Family social group vs male social group0.290.181.680.094
 Laikipia-Samburu vs Amboseli− 1.430.72− 2.000.046
 Maasai Mara vs Amboseli− 3.741.50− 2.490.013
 Tsavo East vs Amboseli− 1.370.40− 3.430.001
 Log (theta)0.380.113.410.001
Zero hurdle model coefficients (binomial with logit link)
 Intercept5.272.691.960.050
 3-month mean NDVI− 56.5629.46− 1.920.055
 Sub-adults and Juveniles vs Adults− 0.710.36− 1.960.050
 Family social group vs male social group0.720.362.010.045
 Laikipia-Samburu vs Amboseli5.042.212.280.023
 Maasai Mara vs Amboseli8.564.381.960.051
 Tsavo East vs Amboseli− 0.250.92− 0.280.783
  1. Abbreviation: SE, standard error