Table 3.
Results from a binomial generalized linear mixed-effects models of WNV antibody detection probabilities in birds sampled in Atlanta, Georgia and Chicago, Illinois (2010–2012)
| Overlap 2010–2012 | |||||
|---|---|---|---|---|---|
| Variable | Estimate | SE | z Value | Pr(>|z|) | |
| Intercept | −6.16 (−8.23, −4.10) | 1.06 | −5.84 | 5.4e−9 | |
| Year: 2011 | 0.79 (0.21, 1.40) | 0.30 | 2.66 | 0.008 | |
| Year: 2012 | 3.15 (2.42, 3.89) | 0.37 | 8.42 | <2e−16 | |
| Age | −12.9 (−17.5, −8.21) | 2.37 | −5.42 | 6.0e−8 | |
| Month | 0.28 (0.05, 0.50) | 0.11 | 2.4 | 0.02 | |
| Month by age | 1.49 (0.90, 2.07) | 0.30 | 5.00 | 5.8e−7 | |
| Random effects | Name | Variance | SD | ||
| Location: species | Intercept | 0.009 | 0.09 | ||
| Species | Intercept | 1.58 | 1.26 | ||
WNV detection was the response term, total monthly trapping effort was included as an intercept offset, month, age, month-by-age interaction, and year (categorical with reference “2010”) were fixed effect terms, and location nested within species was a random effect term. Pr(>|z|) = P value.