Statistics

Abstract

In experimental particle physics as well as in many other fields, it has become increasingly important to analyze data in a manner that extracts the maximum information and takes into account all of the known uncertainties. This article reviews the most important statistical methods used to carry out this task. It begins with an overview of probability, as this forms the basis for quantifying uncertainty. The statistical methods considered include the general framework of statistical tests and parameter estimation, including methods for constructing intervals such as upper limits. Both frequentist and Bayesian approaches are described.