O'Reilly logo
  • James Stevenson thinks this is interesting:

This is a one-sided hypothesis test for a proportion using two samples, and it can be run using a two sample z-test. If I use quota sampling and force the sample size of each group to be the same, or N = N1 = N2, then the test statistic is a z-statistic,


where images is the average of


Here's a good explanation of the formula for calculating sample size for A/B tests.