Figure 3. Association between EEEV exposure and northern cardinal density.

Scatterplot showing the relationship between EEEV exposure risk in chickens during 2009 and 2010 and northern cardinal density at 24 sentinel sites in Walton County, Florida. The estimate for the slope of exposure risk regressed on northern cardinal density was 0.006 with a 95% UCI of [0.0025, 0.0107]. This estimated slope, when an influential observation (indicated by the arrow) was removed from the dataset, was 0.004 [−0.001, 0.009]. The best-fit line from simple linear regression of exposure risk residuals on northern cardinal density are overlaid, with the solid line fit to the full dataset, and the dashed line fit to the dataset that excluded the influential observation.