The futility of small sample sizes for evaluating a binary outcome

This page is being updated from a version on the original website.

You can find an earlier version of this page on my original website.