Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: goodness of MNL fit


From   Maarten buis <[email protected]>
To   [email protected]
Subject   Re: st: goodness of MNL fit
Date   Thu, 2 Oct 2008 14:25:34 +0100 (BST)

--- Chiara Mussida <[email protected]> wrote:
> i'm trying to find a reliable measure of goodness of fit for MNL.
> Fitstat, indeed, shows the well known Mc Fadden R2 but there are
> issues surrounding this indicator.
> Do you know if there are other commands to implement to get such a
> kind of indicator?

There are issues surrounding all indicators of fit. The whole concept
of a measure for the goodness of fit is seriously flawed. First, it
requires a precise definition of goodness of fit, which will always be
arbitrary. Second, what would it mean if you found a bad fitting model?
It may mean that the model is bad, but within the social/economic
sciences I would guess that all models are supposed to have a bad fit:
I hope that we are not entirely determined by our surroundings, though
that is the mechanism I am studying.

So, if I were forced to use one (by advisors/editors/reviewers) I would
pick one, report it, forget it, and focuss on the really important
stuff like interpreting the coefficients or testing the hypotheses of
interest. I certainly would not waste time on looking for "the best"
indicator of fit, as it won't exist. 

-- Maarten

-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room N515

+31 20 5986715

http://home.fsw.vu.nl/m.buis/
-----------------------------------------


      
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index