Kelly Richardson wrote:
> I have 2 categorical variables: race (1=black 0=white) and segregation
> (1=low 2=medium 3=high). I am using the xi command to create the dummy
> variables and then an interaction between the two. So, I have one dummy
> for race and 2 dummies for segregation and 2 dummies for the
> interaction:
>
> race
> segr_med
> segr_high
> race_segr_med
> race_segr_high
>
> Is it possible to get a single p-value for segregation and a single
> p-value for the interaction between race and segregation in addition to
> a p-value for each of the categories? I'd like to look at each of the
> dummies separately as well as the constructs as a whole on a particular
> outcome.
I hope I haven't misinterpreted this, but why not have just one -segr-
variable? After all, it's ordinally measured. That way, you can make the -
not unreasonable, I think - interpretation that for every 1-unit increase
in -segr- (assuming your measurement of it is the best there is) is
associated with Y-unit increase in your dependent variable? (You don't say
what your DV is, but I assume it's continuously measured.) Having just the
one -segr- variable knocks out one of your above interactions, as it's
just race*segregation. If you're happy with this, then you'll be happy
with this interaction. So, you have just one p-value for -segr- and one
p-value for -raceXsegr-. Hope that helps.
CLIVE NICHOLAS |t:
Politics |e: [email protected]
Newcastle University |http://www.ncl.ac.uk/geps
Whereever you go and whatever you do, just remember this. No matter how
many like you, admire you, love you or adore you, the number of people
turning up to your funeral will be largely determined by local weather
conditions.
*
* 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/
