Chapter 3 ran through a fair amount of arithmetic to pin down the coupon effect — group means, standard errors, the standard error of the difference, a confidence interval, a z-statistic, a p-value. There's a single tool that does all of it in one line: linear regression.
Running an A/B Test With Regression
Reanalyze Chapter 2's coupon experiment (control vs. 10% discount) as a regression...