Volume 8, Number 6—June 2002
Research
Drought-Induced Amplification of Saint Louis encephalitis virus, Florida
Jeffrey Shaman*
, Jonathan F. Day†, and Marc Stieglitz*
, Jonathan F. Day†, and Marc Stieglitz*Table 2
Best fit results of the bivariate logistic regression analysis, Floridaa
| Bivariate Best Fits |
||||||||
|---|---|---|---|---|---|---|---|---|
| Site | Antecedent WTD (wks) | Coincident WTD (wks) | Whole model fit p value | Intercept | Antecedent slope | Antecedent slope p value | Coincident slope | Coincident slope p value |
| Site 1 | 12 | 1 | 0.0001 | 29.49 | 26.80 | 0.0005 | -8.98 | 0.0085 |
| Site 2 | 16 | 2 | 0.0001 | 17.83 | 14.13 | 0.0008 | -3.62 | 0.068 |
| Site 3 | 17 | 2 | 0.0001 | 22.11 | 18.64 | 0.0001 | -5.26 | 0.0071 |
| Site 4 | 15 | 1 | 0.0001 | 28.92 | 24.63 | 0.0011 | -7.26 | 0.0096 |
| Site 5 | 15 | 2 | 0.0001 | 21.60 | 23.19 | 0.0002 | -11.37 | 0.0012 |
| All five sites | 17 | 2 | 0.0001 | 19.03 | 18.06 | 0.0001 | -6.21 | 0.0005 |
aThe probability of SLEV transmission incidence is represented as a function of two time lags of weekly averaged modeled water table depth (WTD). Whole model goodness-of-fit and individual parameter estimates were assessed as per the univariate analysis.
