By Claude A Pruneau
Cambridge University Press
Also available at the CERN bookshop
Since the analysis of data from physics experiments is mainly based on statistics, all experimental physicists have to study this discipline at some point in their career. It is common, however, for students not to learn it in a specific advanced university course but in bits and pieces during their studies and subsequent career.
This textbook aims to present all of the basic statistics tools required for data analysis, not only in particle physics but also astronomy and any other area of the physical sciences. It is targeted towards graduate students and young scientists and, since it is not intended as a text for mathematicians or statisticians, detailed proofs of many of the theorems and results presented are left out.
After a philosophical introduction on the scientific method, the text is presented in three parts. In the first, the foundational concepts and methods of probability and statistics are provided, considering both the frequentist and Bayesian interpretations. The second part deals with the basic and most commonly used advanced techniques for measuring particle-production cross-sections, correlation functions and particle identification. Much attention is also given to the notions of statistical and systematic errors, as well as the methods used to unfold or correct data for the instrumental effects associated with measurements. Finally, in the third section, introductory techniques in Monte Carlo simulations are discussed, focusing on their application to experimental data interpretation.
