Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: using inteff after probit w/multi-level categorical variables


From   Lenna Nepomnyaschy <ln77@columbia.edu>
To   statalist@hsphsun2.harvard.edu
Subject   st: using inteff after probit w/multi-level categorical variables
Date   Wed, 3 Oct 2007 09:43:53 -0400 (EDT)

Hi,

I have been using the inteff command to test for interaction effects after probit when there is an interaction b/w a continuous and dummy variable or two dummy variables or two continuous variables.

But, now I have a continuous variable and a 5-level categorical variable and I would like to see the interaction effects for each level, except of course for the omitted one.

For example: depvar=anypay
indepvar1=income
indepvar2=relstat1-relstat5

regular uninteracted model
probit anypay income relstat2-relstat5
*I've omitted the first level of indepvar2

then I create interaction terms for the above
gen increl1=income*relstat1
replace increl2=income*relstat2
replace increl2=income*relstat3
etc.

interaction model:
probit anypay income relstat2-relstat5 increl2-increl5

the inteff command is not reading this model properly because there are two many variables and they are not in the right positions, I think.

Does anyone have a better idea? I want to see the interaction effect for each level of the relstat variable.

thank you,
Lenna



Lenna Nepomnyaschy, PhD
Associate Research Scientist
Columbia University School of Social Work
1255 Amsterdam Ave, room 718
New York, NY 10027
Phone: (212) 851-2379
Fax: (212) 851-2206
*
* 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/




© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index