1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Textbooks on Error Analysis

  1. Mar 12, 2012 #1
  2. jcsd
  3. Mar 12, 2012 #2


    User Avatar
    Science Advisor
    Gold Member

    Bevington is the standard book on error analysis for physics students and physicists. It's well written and relatively easy to understand. I'd definitely start there.
  4. Mar 13, 2012 #3

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor

  5. Mar 13, 2012 #4
    Since I don't know very much about either, why this over Bevington?
    That looks like same publisher as the Taylor book on Mechanics, assuming they're the same author I would think that Taylor's book should be very clearly written?

    After viewing the introductions/prefaces for both it would appear that the Taylor book is truly geared for a first viewing of the material. That's definitely a good thing, but I may need some (slightly) more advanced material after I finish. Do you have a recommendation for a reference or slightly more advanced book to follow Taylor?

    Sorry for asking so many questions ^^;
    Last edited: Mar 13, 2012
  6. Mar 13, 2012 #5
    This is what was said to me:

    "As far as data analysis goes, beyond the basics (mean, sigma, chi-squared fits) it would be good for you to understand maximum likelihood fits (particularly multi-dimensional ML fits). Also, we probably will want to explore using multi-variate discriminators."

    Since I took AP Stat back in high school, I can't really say I've had Stats :P because that class was really poorly put together, so I'm definitely going to recover the basics.

    Now, for Bevington [3rd edition] I'm looking at covering:
    Chapter 1 - Uncertainties in Measurements
    Chapter 2 - Probability Distributions
    Chapter 3 - Error Analysis
    Chapter 4 - Estimates of Means and Errors
    Should I cover Chapter 5 - Chapter 9?
    I'm not sure what I will need as prereq to beginning Chapter 10 on Maximum Likelihood...and I'm guessing I'll need Chapter 11 - Testing the Fit

    For Taylor [2nd Edition] I'm looking at the following:
    (from Part I)
    Chapter 1. Prelinary Description of Error Analysis
    Chapter 2. How to Report and Use Uncertainties
    Chapter 3. Propagation of Uncertainties
    Chapter 4. Statistical Analysis of Random Uncertainties
    Chapter 5. The Normal Distribution
    (from Part II)
    Chapter 6. Rejection of Data
    Chapter 7. Weighted Averages
    Chapter 12. The Chi-Squared Test for a Distribution
    I'm not sure what parts of Chapters 8 - Chapter 11 I will need.

    The work I'll be doing [though I don't know exactly what yet] will be dealing with particle data, if this helps in determining what I need to know. I'm learning to work with a various packages in ROOT like RooFit and TVMA, but again I'm not sure if that helps for context much since I'm not very far into this yet!

    I am not entirely sure that Bevington (I doubt Taylor has the material) has the last two things he mentioned: Multidimensional MLs and Multivariate Discriminators, so it would be great to know if they do or where I can learn this material. To be honest, I don't even know what is meant by the last term and Google did not give me an immediate answer.

    Any recommendations on what to cover from where? :P I've got three whole months to work on this stuff, so I should be okay double covering a bit, right?
  7. Mar 13, 2012 #6
    Taylors error analysis book may not be super advanced but Taylor is a fantastic author in my opinion. Grab his book!
  8. Mar 13, 2012 #7
    This is the feeling I've been getting in general, I have a copy on hand ( :!!) University Libraries) and it feels sort of like his Mechanics book.

    Do you have any recommendations for an advanced book to follow his?
  9. Mar 13, 2012 #8

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor

    I have a few texts on error analysis, for whatever reason Taylor is the easiest one for me to use- definitely check a few different ones out of your library, YMMV :)
  10. Mar 16, 2012 #9
    I know this isn't a popular subject, but I'd like to bump this just to see if anyone has any other input.

    Someone recommended this to me as well:

    So I'm looking at a combined approach reading through that paper, and working through bits of Bevington and Taylor together.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook