st: RE: comparing nested models

[Susan Bondy <sue.bondy@utoronto.ca>]

Look up the lrtest command.

The two models are nested and can be compared using a likelihood ratio test (LRT) if the fully-spelled-out set of beta coefficients in the model are nested (one is a reduced set of the other) and the observations are identical and data are independent (e.g., not survey data).  If the five categories are the 10 categories collapsed in pairs, for one example, you can compare using a LRT.  

If models being compared are on the same observations but you are working with non-independent data or the two models are not nested (i.e., same number of beta coefficients) then there are measures of goodness of fit to allow non-nested models to be compared and people have suggested means to test critical levels of improvement of fit using these measures.  Look up AIC BIC and VIC in your favourite regression texts.

-Sue



-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nikolas Asasa
Sent: February 3, 2011 11:09 AM
To: statalist@hsphsun2.harvard.edu
Subject: st: comparing nested models

Dear All, 
I have two models aiming to estimate effect of age on wage. In the first model age is recorded in 5 categories. In the second model age is recorded in 10 categories.
The question is how to compare goodness of fit of two models? 

In principle I would apply F-test, but I am not sure whether "test" procedure in Stata allows me to do it? Should I use ANOVA instead?

Thank you for your replies.
Anton


      
*
*   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/

*
*   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/