2.7 Confidence Intervals

A confidence interval, like the standard error, is a way to estimate the reliability of a sample statistic.

For example, a 95% CI is an interval that would contain the central 95% of values for the sample statistic, if the sampling experiment were done a very large number of times.

It’s generally not practical to actually sample that many times, so… bootstrap!

For comparison, plus or minus one SE gives a CI of about 68%.