Skip to main content
. 2017 May 5;12(5):e0176297. doi: 10.1371/journal.pone.0176297

Table 3. Maximum likelihood estimates of effects in the best-approximating model of AIV infection in BC.

Variable Term MLE SE 95% CI
Intercept -2.033 0.246 -2.515, -1.551
Sex (ref = Female) Male 0.354 0.101 0.156, 0.553
Species (ref = MALL) AGWT -0.975 0.363 -1.687, -0.264
AMWI -0.463 0.498 -1.440, 0.513
BWTE -0.679 0.3500 -1.364, 0.007
NOPI -1.054 0.437 -1.911, -0.197
Year (ref = 2005) 2006 -1.025 0.126 -1.273, -0.778
2007 -1.593 0.158 -1.903, -1.283
2008 -1.228 0.235 -1.688, -0.768
2010 -0.762 0.257 -1.267, -0.258
Sampling location (ref = Alaksen) Minnie Lake 2.127 0.235 1.666, 2.587
Nicola Lake 2.565 0.2300 2.114, 3.015

The model was fitted as a generalized linear model with a sample size of 2281. All values shown are on the logit scale. Exponentiation of estimates provides the odds ratio interpretation of effect size. MLE is the maximum likelihood estimate of the parameter, SE is the standard error of the estimate, and 2.5% and 97.5% values define the 95% confidence interval around the MLE.

Abbreviations: MALL: A. platyrhynchos, AGWT: A. carolinensis, AMWI: A. americana, BWTE: A. discors, NOPI: A. acuta