Re: st: interaction terms in a Cox model

[Steven Samuels <ssamuels@albany.edu>]

In general, failure to reject the null hypothesis does not mean that  
it is true.  Tests for interactions notoriously have low power.  If  
there is one strong interaction (one age category, say) it will be  
masked and an omnibus test could be insignificant.  That said, the  
size of the z statistics in your case indicates that the interactions  
are weak at best.  You should, however, look at the full two-way  
table of HR's (age group x gender) to satisfy yourself that this is  
the case.  That is create a 6 category age-gender combination  
variable ( "egen age_gender= group(agecat gender) will do this.,

Steven

On Apr 14, 2007, at 9:06 PM, Ashwin Ananthakrishnan wrote:


> Hi,
>
> Since the P >0.05 for the null hypothesis that the
> interaction terms are equal to zero, is it safe to
> conclude that there is no significant interaction
> between race and gender?
>
> *   http://www.ats.ucla.edu/stat/stata/
*
*   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/