Skip to main content
. 2012 Sep 25;7(9):e45789. doi: 10.1371/journal.pone.0045789

Table 2. Model selection results using the information theoretic approach1.

Predicting Leucocytozoon infection
Model AICc ΔAICc AICc Weights Model Likelihood Parameters Deviance R-squared
{ψ area, ρ hematozoa prevalence} 436.4206 0 0.82122 1 4 428.2882
{ψ wind, ρ hematozoa prevalence} 439.4699 3.0493 0.17878 0.2177 3 433.3907 0.949981718
{ψ migration duration, ρ hematozoa prevalence} 500.5873 64.1667 0 0 3 494.5081 0.350866118
{ψ migration distance, ρ hematozoa prevalence} 504.4218 68.0012 0 0 3 498.3426 0.313277661
{ψ temperature, ρ hematozoa prevalence} 516.8494 80.4288 0 0 3 510.7702 0.191453613
{ψ, ρ hematozoa prevalence} 534.3404 97.9198 0 0 2 530.3009
Predicting Haemoproteus / Parahaemoproteus / Plasmodium infection
Model AICc ΔAICc AICc Weights Model Likelihood Parameters Deviance R-squared
{ψ area, ρ hematozoa prevalence} 277.4747 0 0.95578 1 4 269.3423
{ψ wind, ρ hematozoa prevalence} 283.6705 6.1958 0.04315 0.0451 3 277.5913 0.938686787
{ψ temperature, ρ hematozoa prevalence} 291.0593 13.5846 0.00107 0.0011 3 284.9801 0.883767273
{ψ migration duration, ρ hematozoa prevalence} 405.5536 128.0789 0 0 3 399.4744 0.032753401
{ψ migration distance, ρ hematozoa prevalence} 407.21 129.7353 0 0 3 401.1308 0.020441702
{ψ, ρ hematozoa prevalence} 407.9204 130.4457 0 0 2 403.881
1

Unresolved samples considered as negative for infection.