## Propagation of Error and Confidence Intervals for Birds?

Science-based work isn’t always fun. Often it can be tedious and time-consuming, so why do it? Well, because nothing is ever 100 percent, ever! So if your data looks perfect but your confidence interval is large, then you just got lucky collecting that set of data and it is not truly correlative. Confused yet? What if your doctor told you that a medication is 75 percent effective, give or take 20 percents of freedom. So that medication could be anywhere from 55 to 95 percent effective; in other words, the doctor and the drug company have a very vague idea of how effective it is. On the other hand, if a doctor told you that Â a medication was 95 percent effective with 1 percent of freedom, then that medication is 94 to 96 percent effective. That confidence interval makes a significant difference in how much that data can be correlated and interpreted. Knowing how to correctly calculate those intervals is a PAIN, but it is necessary. I’m no expert yet, but I understand the concept and have learned how to incorporate this type of analysis into the data that I have produced from personal research. This, consequently, applies to migratory birds. In doing surveys to count birds, there is over-counting and many factors that contribute to uncertainty in bird numbers. But by understanding the methods of counting you can shrink your confidence interval for the bird counts you do, which makes your survey data more valuable!

##### Edgar Escobedo
edgareescobedo@gmail.com