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: Number of degrees of freedom


From   Richard Goldstein <richgold@ix.netcom.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Number of degrees of freedom
Date   Sat, 27 Feb 2010 09:01:26 -0500

it appears that, although you say that A and B are categorical, when you
form your interaction you are treating them as continuous

you don't show your commands (as the FAQ recommends)

if you are using version 11, I recommend you use "factor variables" and
that will give you what you are looking for (if you are using an earlier
version, use the -xi- prefix)

Rich

Nikolaos Pandis wrote:
> Hi to all.
> 
> I would like to ask:
> 
> Let us say we run a Cox model with 2 predictors A and B, where A has 4 levels and B has 2 levels.
> Then you run the same model plus the A*B interaction term and perform an 
> -lrtest- to check if the interaction is significant.
> 
> How many degrees of freedom would this test have? Is it 3? 
> Why? 
> Is it because 3 extra parameter are included in the second model as we look at the output? 
> Do we count the difference in the number of parameters in the output between the 2 models or the difference in the number of parameters in the command line? The latter will indicate only 1 parameter difference i.e the intarction term
> 
> Many thanks,
> 
> Nick
*
*   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