2.6 The Bootstrap Solution

If our sample statistic is the mean⁵, we can estimate the standard error from a single sample.

Otherwise, to measure the standard error of our sample statistic, we would have to repeat our sampling process many times to be able to calculate the standard deviation of the sampling distribution. (Of course, if we actually did this, our calculated standard error would be pretty useless, because we’d have a better estimate of the statistic from the combined samples.)

In practice, we can use a bootstrap.

---

5. or other statistic for which a theoretical approximation to the standard error has been derived↩︎