[Ifeffit] statistics

Matt Newville newville at cars.uchicago.edu
Tue Aug 24 10:24:49 CDT 2004


> When it comes down to compare fits using different models, which is 
> the relevant parameter? Chi square (and reduced) or R-factor? I have 
> fits with smaller chi square and larger Rfactor and viceversa...

This is definitely a FAQ, so I put the answer at

> Is it possible that a rationale would be that when the models to 
> compare have the same number of atoms these parameters go in 
> parallel? And that when the number of atoms is different one must 
> rely more on chi square rather than R-factor (or viceversa)?

I'm not sure I understnad this:  I'd say the number of atoms or 
paths, or anything else that describes the complexity of the 
model, is NOT important. What's important is how well the 
specified variables cause the model to match the data. 


More information about the Ifeffit mailing list