Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Model goodness-of-fit


From   "Subhash Pokhrel" <Subhash.Pokhrel@urz.uni-heidelberg.de>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Model goodness-of-fit
Date   Wed, 15 Sep 2004 11:23:41 +0200

Dear List members,

I have two continuous variables and eight dummy variables in a logistic
regression model (n= about 17,500). I tried to test the model goodness of
fit with Hosmer Lemeshow test which gave me large chi2 value with very small
p value, showing the lack of fit. If I drop one of the two continuous
variables from the model, I get much smaller Chi2 with large p values. Since
this is the most important variable I would like have in the model, I would
appreciate receiving your suggestion as to how I could solve this problem.
Is there any other tests available in Stata? Surprisingly, while searching
the net I came across a very big criticism of using Hosmer-Lemeshow test.
Thanks,
Subhash.



*
*   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