How to use the NIS data
The National Immunization Survey (NIS) has a sample size of about 15,000 children with adequate provider data. However, the sample is not the entire population, and thus sample estimates and population values are likely to be different. The magnitude of the likely difference is quantified through the 95% confidence interval. For example, in 2014, the vaccination coverage estimate for 3 doses of poliovirus vaccine in Alabama was 90.5% +/- 6.0%. This means that the true coverage was probably between 84.5% and 96.5%. When comparing two estimates (e.g., between states or between years), an overlap in the confidence intervals indicates that the observed difference might be due to chance.