I'm wondering if there is a textbook that describes the handiwork of a particle physics analysis. There are a bunch of books about theory, about the experimental aspects like detectors, and about statistical methods, but I haven't seen one that focused on the actual work of an analyzer.

I mean stuff like:

  • How to conduct a search vs. a measurement
  • What backgrounds to consider for a final state, and how to model them
  • What distributions to look at -- pT of various objects or MET are obvious, but there are things like transverse masses, $\cos \theta^*$, aplanarity, the Florida variable (seriously), or ptBalance, that you first have to know about
  • How to choose signal regions, sidebands
  • How to determine systematic errors concretely
  • How to perform a cutflow
  • How to do tag-and-probe
  • How to perform a multivariate analysis, and how to choose btw. techniques
  • How to set limits, and how to perform combinations

Most of this stuff is covered in various books, but mostly from a different perspective. For example, its nice to know how to solve an integral with Monte Carlo methods, and what the factorization theorem does. But for someone working on an analysis its more useful to know what the main differences between MC generators are, and how to deal with negative event weights in various situations. Similarly, there are a couple of textbooks about statistics for high energy physics, but those I've found tend to focus on derivations, instead of practical issues of the analysis.

Does anyone know of a book that fits my description?

(Note I don't believe this fits very well under the current book policy. Resource recommendation questions tend to be fairly broad, and thus the answers have to be very descriptive. In this case, the question is already descriptive, so a brief answer would also be OK, even a negative answer (a la "I've been a expert for 20 years and can say for certain that such a book doesn't exist").)

  • 3
    $\begingroup$ Don't know of one and would expect that I have heard of it is there was one. I was a postdoc for four years in a High Energy Physics group that offers REUs as well as trains grad students and my bosses there did not know of such a thing. $\endgroup$ – dmckee Dec 6 '13 at 16:57
  • 1
    $\begingroup$ @dmckee how about the lecture notes of Tilman Plehn? I'm not an experimentailist myself, but I liked them arxiv.org/abs/0910.4182 $\endgroup$ – innisfree Dec 18 '13 at 11:33
  • $\begingroup$ @innisfree Interesting. I hadn't seen that before. $\endgroup$ – dmckee Dec 18 '13 at 15:30

This could be probably the closest one that I know:

Data Analysis in High Energy Physics: A Practical Guide to Statistical Methods Olaf Behnke (Editor), Kevin Kroninger (Editor), Gregory Schott (Editor), Thomas Schorner-Sadenius (Editor) ISBN: 978-3-527-41058-3 http://eu.wiley.com/WileyCDA/WileyTitle/productCd-3527410589.html

Especially Chapter 11 is explaining an analysis from beginning to the end. Here you can find the content of Chapter 11 from the book:

Analysis Walk-Throughs by Aart Heijboer and Ivo van Vulpen

  • Introduction
  • Search for a Z 0 Boson Decaying into Muons
  • Counting Experiment
  • Quantifying the Sensitivity: p-Values and Significance
  • Optimising the Mass Window
  • Estimating the Background from Data Using Sidebands
  • Scanning over the Full Dimuon Mass Range: The ‘Look-Elsewhere Effect’
  • Profile Likelihood Ratio Analysis

  • Profile Likelihood Test Statistic

  • Properties of the Test Statistic Distributions for the b-only and s C b Hypotheses
  • Rules for Discovery and Exclusion
  • Results from Data
  • Probing the Sensitivity Limits: Enhanced Luminosity and Signal Cross Sections
  • Scanning the Full Mass Region
  • Measurement
  • Introduction
  • Unbinned Likelihood
  • Likelihood Ingredients
  • Extracting a Measurement in the Presence of Nuisance Parameters
  • Mass Measurement
  • Testing for Bias and Coverage
  • Systematic Uncertainties
  • Constraints and Combining Measurements
  • $\begingroup$ Hi Bora, we require answers to contain at the minimum, enough information that readers don't need to follow the link to get any value. Please include a summary of the information you've linked to in your answer. $\endgroup$ – Brandon Enright Feb 5 '16 at 23:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.