Objections to Frequentism

(Updated 6/1/19)


Hi. I'm a use whatever works type of statistician: frequentism, Bayesian, SAS, Python, whatever. However, after reading and hearing (and collecting) many arguments against frequentism, some of them contradicting, and most of them rehashed year after year, I decided to write this article.

Frequentism is roughly defined as:

Criticisms of Frequentism:

Responses to Criticisms of Frequentism:

Frequency definition of probability

Frequentist concepts like hypothesis testing and p-values are too difficult to teach.

Frequentism relies on data you didn't observe.

Frequentism is bad for science.

Fisher dislike or envy

Frequentism is used to p-hack.

Frequentism is responsible for the "replication crisis".

Just look at all those counterexamples to frequentism!

  • Counterexamples, paradoxes, or issues in Bayesian probability and statistics
  • Frequentism is really just a special case of Bayesian.

    But everything is subjective anyway!

    What is your background?

    Thanks for reading and sending in comments/corrections!

    Comment from reader:

    I got many answers in the frequentist vs Bayesian thread, but this pro-frequentist page is really superb!

    If you enjoyed any of my content, please consider supporting it in a variety of ways: