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: problem with factor variables in RE probit
From
Richard Williams <[email protected]>
To
[email protected]
Subject
Re: st: problem with factor variables in RE probit
Date
Tue, 27 Sep 2011 10:29:32 -0400
At 07:41 AM 9/27/2011, Liberini, Federica wrote:
Dear listers,
I am having problems using the factor variables in a RE probit model.
My "group" variable is a categorical that can take up to 3 different
values. My "X" variable is a continuous.
What I did was to compare two regressions:
tab group, gen(group)
gen gr1_x=group1*x
gen gr2_x=group2*x
xtprobit y x group1 group2 gr1_x gr2_x xbar zbar, re
compared to
xtprobit y b3.group##c.x xbar zbar, re
(the xbar and zbar variables are for the RE specification)
Now, before running the two regressions I made sure that the virtual
variables generated by b3.group##c.x were identical to the varlist x
group1 group2 gr1_x gr2_x . I did this by typing
list x group1 group2 gr1_x gr2_x b3.group##c.x in 1/100
and they were!
So why is that the two regression command return different outcomes?
You don't say how the outcomes are different. Are the fits identical
but the parameterizations are different? Personally, I tend to avoid
the ## notation and just try to write things the same way I would if
I had separate variables. That way I feel more sure of what I am
asking for. So, I would rewrite your first command as
xtprobit y x b3.group b3.group#c.x xbar zbar, re
If that doesn't work then provide more detail as to how the results
are different.
-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME: (574)289-5227
EMAIL: [email protected]
WWW: http://www.nd.edu/~rwilliam
*
* 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/