The Means of Randomization

Episode 198 · May 28th, 2019 · 1 hr 33 mins

About this Episode

How would you feel if you found out you were unwittingly the subject of an experiment testing two alternatives? You got A, and another group got B. Many people object to this. But what if neither A nor B was at all objectionable and in fact each is served up at many other places unilaterally and without reason for preferring one to the other? Why should we object to being randomly given A or B for the purpose of testing, when we would not object to having either uniformly and arbitrarily imposed? We are joined again by Michelle Meyer to discuss this problem, made famous recently by Facebook and other A/B testing entrepreneurs.