Thanks for your reply, Christer. I guess I was not exact enough.
What I have are price and characteristics observations for many
different products, similar to the example-data of 1978 automobiles,
plus information to which of the five segments ("groups") of the
market they belong. I want to perfom a hedonic regression for every
segment, with restrictions across equations on some, but not all of
the coefficients. I thought therefore I needed to estimate some
system of related equations instead of five completely unrelated
regressions...
Interacting the variables does not seem to be an option here as I
have a lot of explanatory variables, especially (firm and other)
dummies - I already had to reduce my initial model for -matsize-
considerations.
Jonathan
> Jonathan wrote:
>
> I want to perform a SUR with 5 identical equations, but for 5
> different groups of observations in my data, and make some
> restrictions and tests on coefficients across equations. My problem
> is: I don't know how to tell Stata that these 5 equations are not the
> same, although they all contain the same variables.
>
> Are you sure that you are adopting the right strategy?
>
> As I understand it, SUR is the thing to do when you have 2 (or more)
> correlated dependent variables but the independet variables are similar
> across the equations. The problem you have, as I see it, can be solved more
> easily by estimating 5 different equations (one for each group) and compare
> the magnitude of the different independent variables in the traditional
> way.
>
> Or you can create product terms (interaction variables) involving the
> 5-group variable and the remaining independent variables in a more
> comprehensive equation...
>
> Well, that's my comment...
>
> Christer
>
> Christer Thrane, dr.polit.
> Associate Professor, Sociology
> Faculty of Tourism and Applied Social Science
> Lillehammer College
> Gudbrandsdalsv. 350
> 2626 Lillehammer, Norway
> + 47 61 28 81 70 (fax)
> + 47 61 28 82 47 (phone, work)
> + 47 61 25 68 44 (phone, home)
> + 47 92 29 54 39 (mobile)
> E-mail, work: [email protected]
> E-mail, home: [email protected]
>
>
> Dear Listeners,
>
> I want to perform a SUR with 5 identical equations, but for 5
> different groups of observations in my data, and make some
> restrictions and tests on coefficients across equations. My problem
> is: I don't know how to tell Stata that these 5 equations are not the
> same, although they all contain the same variables.
>
> I generated distinct variables for every group, like p1-p5 for the
> endogenous and x1-x5 for one of the exogenous variables by saying:
> . gen p1=.
> . replace p1=p if group==1
> and so on.
>
> But to my command
> . sureg (p1 x1) (p2 x2) ? (p5 x5)
> Stata replied:
> insufficient observations, r(2001)
> although every group has hundreds or thousands of observations.
>
> In the archive, I found a short response by Kit Baum pointing to a
> longer discussion that took place in early 2002 on a this issue, but
> that I couldn't find.
>
> I'd be very grateful if someone could give me a more detailed
> reference of that earlier discussion, or some other clues.
>
> I use Stata 7.
>
> Kind regards
> Jonathan
>
> *
> * 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/
*
* 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/
