Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Observations differ


From   Nick Cox <njcoxstata@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Observations differ
Date   Sun, 23 Sep 2012 13:18:39 +0100

I imagine one model is using fewer observations because of missing
values in one or more predictors, a problem, which doesn't bite for
the other model. Or there could be problems if predictors are
constant.

A strategy includes

0. Look carefully at the structure of missing values. See e.g.
-misstable-, -nmissing- (SJ).

1. Identify the model using fewer observations.

2. Define

gen used = e(sample)

after fitting that model.

3. Fit the other model using

... if used

4. Compare fits if possible.

Nick

On Sun, Sep 23, 2012 at 1:02 PM, Shittu, Aminu <ameen_vet@yahoo.com> wrote:

> I am trying to compare, using likelihood ratio test, between the full and reduced model in a random effect logistic regression; but the output is saying "observations differ: 3356 vs. 3583." How can I handle this problem, please.
*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index