Telling the truth with statistics

by F. James (CERN-EP)

Thursday 13 Jan 2000, 11:00 → 14:15 Europe/Zurich

500/1-001 - Main Auditorium (CERN)

Lecture 4 How to tell the truth with Statistics? Lying with statistics is easy. A greater challenge is to use statistics to communicate an accurate and objective summary of given experimental data. We will follow both the Bayesian and classical (frequents) approaches to this problem, and will see that in both cases the key to getting the right answers will be to ask meaningful questions. A working knowledge of elementary statistics will be assumed, so that we can concentrate on exact methods and questions of interpretation.