It is Time to Stop Teaching Frequentism to Non-statisticians
arxiv.org
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.
  • witty_username@feddit.nlEnglish
    1·
    11 hours ago

    Although I’ll admit that I just skimmed the article, it seems like the author is simply stating and illustrating his opinion without supporting it with any data or stats. Bayesian stats are easier than frequentist stats? For what kind of data? In what kind of research setting? In what kind of discipline? For what kind of student?

    • FrenziedFelidFanatic@pawb.socialEnglish
      1·
      11 hours ago

      It’s basically a philosophy paper—epistemology, to be specific.

      Frequentism and Bayesianism are philosophies of understanding probability and statistics, and this paper argues that Bayesianism provides a more clear understanding of truth.

      To summarize their main points: P-values and confidence intervals and their ilk are confusing to the uninitiated and initiated alike, so their use should be discouraged. Undergraduates shouldn’t even be taught them since they confuse most people, causing public harm. All frequentist statistics at this level should be purged from education and replaced with Bayesian statistics.