Statalist


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

Re: st: Hosmer-Lemeshow test with clogit


From   "Peter Dijkstra" <ptdijkstra@gmx.net>
To   statalist@hsphsun2.harvard.edu, statalist@hsphsun2.harvard.edu
Subject   Re: st: Hosmer-Lemeshow test with clogit
Date   Wed, 06 Feb 2008 15:47:15 +0100

That's a good idea!
I have a 4-option model, then I have the problem that a threshold of 25% nor the maximum probability are perfect as threshold. However, what if I'd like to select the maximum value? I use
    by each_individual: gen correctly_predicted = (probability == max(probability))
But this ain't working: what do I do wrong?
Cheers, Peter

-------- Original-Nachricht --------
> Datum: Wed, 06 Feb 2008 14:59:58 +0100
> Von: Thomas Gschwend <gschwend@uni-mannheim.de>
> An: statalist@hsphsun2.harvard.edu
> Betreff: Re: st: Hosmer-Lemeshow test with clogit

> I successfully convinced reviewers with the "percentage correctly 
> predicted" using  -predict- and then cross-tabulate the predicted values 
> (assuming a threshold of .5) with the values of your dependent variable.
> Thomas
> 
> Peter Dijkstra schrieb:
> > After a normal logit model you can use the Hosmer-Lemeshow test to check
> the goodness-of-fit. However, I estimated a conditional logit (clogit)
> model and am unable to use lfit (I still use Stata8, in release 9 this
> corresponds to "estat gof"). Is there a similar test available somewhere?
> > Cheers, Peter
-- 
GMX FreeMail: 1 GB Postfach, 5 E-Mail-Adressen, 10 Free SMS.
Alle Infos und kostenlose Anmeldung: http://www.gmx.net/de/go/freemail
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index