When comparing means of two non-negative random variables, we are often interested in the ratio of two means since the ratio does not depend on the scale of underlying distributions.
If we know when a potential change might possibly happen in a stochastic process then we can set a sequential test to continuously monitor whether we can tell the change
In this follow-up post, we explain another advantage of sequential hypothesis testing: Flexibility and safety. Please visit the previous post for an introduction to Wald’s
In this and a follow-up posts, we explain two main advantages of sequential hypothesis testing methods compared to standard tests based on fixed sample size. Sample efficiency in practice As