This is a deck I’ve been using over the past few months to teach the product, program, business development, and marketing folks at RunKeeper about some of the perils and pitfalls of A/B testing. I intended it to be a good first step for people interested in testing but without much background in statistics. Enjoy!
(By the way, if you’re comfortable with IPython and would like to play around with the sample size calculations described on Slide 12, here is the IPython notebook that I put together to run the numbers.)
