Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

Re: st: interaction effects vs. interaction coefficients using contrast?


From   Colleen Nugent <[email protected]>
To   [email protected]
Subject   Re: st: interaction effects vs. interaction coefficients using contrast?
Date   Mon, 06 Feb 2012 15:10:26 -0500 (EST)

Thanks again, Maarten.

I guess I can't really make that kind of alteration that you would make for your Stata tip example.  I have a three-category by three-category interaction and I'm interested in several different comparisons for different reasons.  So it looks like margins with lincom might be the simplest way to do that.

So my second question then is:  I tried the margins syntax you suggest in the Stata tip, but I get the error "option over() not allowed r(198)."  I am assuming this is because -over- treats each group as its own subpopulation, but I am already specifying a subpopulation that is my analytic subsample using svy, subpop().  Instead I have tried "margins varA#varB, [options]" and "margins varA, at(varB=(1 2 3)) [other options]" and I get slightly different results.  I'm not quite sure which is the best substitute for the over option (or is there some other?).  

Many thanks for your time and expertise!
Colleen


----- Original Message -----
From: "Colleen Nugent" <[email protected]>
To: [email protected]
Sent: Monday, February 6, 2012 10:29:26 AM
Subject: Re: st: interaction effects vs. interaction coefficients using contrast?

Thank you, Maarten.

Following the example you gave in the Stata tip, if I wanted to know whether black women with a college degree had significantly different odds than white women with a college degree, the best way to do it is to use margins and then lincom to make this comparison?  

Also, I tried the margins syntax you suggest, but I get the error "option over() not allowed r(198)."  I am assuming this is because -over- treats each group as its own subpopulation, but I am already specifying a subpopulation that is my analytic sample using svy, subpop().  Instead I have tried "margins varA#varB, [options]" and "margins varA, at(varB=(1 2 3)) [other options]" and I get slightly different results.  I'm not quite sure which is the best substitute for the over option (or is there some other?).  (I'm brand new to Stata 12, margins, and even the factor variable notation as I used Stata 10 previously.)

Thanks for your help!
*
*   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 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index