A Quick Explanation on Why 95% Confidence Intervals Get Smaller (For Non-Epidemiologists)

I’ve been tutoring some students in online MPH programs for a few weeks. We started talking about 95% confidence intervals. All you need to know when reading a confidence interval is that it is used for telling you (the reader) where we (the researcher) are 95% confident that the true measure is located.

Here, let me explain…

This is why polls have confidence intervals but the general election doesn’t. Though, as a friend stated, maybe Florida should have a confidence intervals.

Of course, 95% Confidence Intervals don’t tell the whole story. You have to know what the study is getting at, how it is designed, and what measures you’re comparing. And you have to decide what kind of a confidence interval you can live with.