Lectures on the Foundations of Applied Statistical Inference
WORKING PAPER
(3.6 MB)
This book aims to make explicit exactly what implicit assumptions about the informational state of the observer support frequentist analysis.
The most coherent way to understand the logic of probabilistic and statistical arguments is to start from the Laplacian point of view that explicitly puts an observer in possession of explicitly defined information in the center of the theory. It is possible from this starting point to give a satisfactory and transparent account of frequentist methods and results. Basically, I argue that frequentist methods are Laplacian methods derived from particular assumptions about the informational state of the observer.