Bayesian vs Frequentist Perspectives

An Excellent Explanation

In #articles

Bayes

Jake VanderPlas has what I believe is the clearest, and most thorough, take-down of the frequentist approach to statistical inference. It's in four parts, so far. The third part, in my opinion, is the clearest example of the differences in the approaches, demonstrating the superiority of the Bayesian approach, and why frequentist statistics is wrong in science. The four parts are:

  1. Frequentism and Bayesianism: A Practical Introduction
  2. Frequentism and Bayesianism II: When Results Differ
  3. Frequentism and Bayesianism III: Confidence, Credibility, and why Frequentism and Science do not Mix
  4. Frequentism and Bayesianism IV: How to be a Bayesian in Python

One of the reasons I particularly like this series is that the examples are explored with simple Python code, which makes it completely obvious what is going on. Computer simulations force one to put all of the assumptions down explicitly, and to see what the practical differences are.

Once you read this, and convert to Bayesianism, you may want to read my book too which takes this perspective without focusing on the debate.