It is Time to Stop Teaching Frequentism to Non-statisticians

TL;DR

The paper advocates replacing frequentist statistics with Bayesian methods in undergraduate education to reduce confusion and overconfidence among non-statisticians.

Contribution

It proposes a paradigm shift in teaching statistics, emphasizing Bayesian methods over frequentist approaches for clarity and better understanding.

Findings

Bayesian methods can decrease misconceptions among students.

Switching to Bayesian teaching reduces over-certainty in statistical interpretation.

The paper highlights the benefits of Bayesian over frequentist approaches in education.

Abstract

We should cease teaching frequentist statistics to undergraduates and switch to Bayes. Doing so will reduce the amount of confusion and over-certainty rife among users of statistics.