What are some recommended textbooks for error analysis in scientific research?

[Elwin.Martin]

I was recently told I needed to review some material and they recommended I look at Bevington: http://books.google.com/books/about/Data_reduction_and_error_analysis_for_th.html?id=0poQAQAAIAAJ
I haven't had a chance to look at it yet, but I was wondering if anyone had any other recommendations? My background is mostly in Analysis and Algebra :S So while I've had a course in Discrete, this is a little unfamiliar in general.

 

[marcusl]

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.

 

[Andy Resnick]

My favorite book, *by far*, is Taylor's:

http://books.google.com/books?id=gi...wAA#v=onepage&q=taylor error analysis&f=false

 

[Elwin.Martin]

My favorite book, *by far*, is Taylor's:

http://books.google.com/books?id=gi...wAA#v=onepage&q=taylor error analysis&f=false

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?

Edit:
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 ^^;

 

[Elwin.Martin]

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?

 

[Jorriss]

Taylors error analysis book may not be super advanced but Taylor is a fantastic author in my opinion. Grab his book!

 

[Elwin.Martin]

Taylors error analysis book may not be super advanced but Taylor is a fantastic author in my opinion. Grab his book!

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?

 

[Andy Resnick]

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?

Edit:
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 ^^;

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 :)

 

[Elwin.Martin]

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:
http://arxiv.org/pdf/physics/9711021v2.pdf

So I'm looking at a combined approach reading through that paper, and working through bits of Bevington and Taylor together.