If you think that pure frequentist stats are only a starting point and that Bayes was right (and I do) and if you struggle to remember Bayes’ Theorem represented as a set of probabilities (and I do) then you’ll really appreciate the messages laid out very clearly in Gerd Gigerenzer’s book “Reckoning with Risk”.

He makes and supports a number of compelling points:

  • We live in a world of uncertainty – get used to it.
  • In such a world it is important to communicate effectively (and to understand accurately so that you don’t get exploited) what these uncertainties really mean.
  • There are a number of different ways of representing (to understand and to communicate) uncertainty – and natural frequencies have a natural and intuitive immediacy that gives them a huge advantage
  • And, by the way, Bayes becomes obvious when you use natural frequencies.

Brilliant book, just finished it and highly recommend it.