Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: seemingly unrelated regression


From   John Antonakis <john.antonakis@unil.ch>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: seemingly unrelated regression
Date   Fri, 04 Jun 2010 15:22:02 +0200

You need to stack the models, i.e., estimate the models separately for each group. I don't know what you reviewer is saying, but I can see how what they suggest needs to be done with SUR.

So, take a look a -help suest-. With this command, you can estimate the regression model separately for each group and then combine the results with a robust or cluster robust variance estimator like this:

y = x a, if group == 1
est store one
y = x a, if group == 2
est store two
suest one two, cluster (id)

Then you can do a "Chow test" or what have you across the equations. This is basically testing a moderation hypothesis; see the help file or manual for details.

HTH,
John.

____________________________________________________

Prof. John Antonakis, Associate Dean Faculty of Business and Economics
Department of Organizational Behavior
University of Lausanne
Internef #618
CH-1015 Lausanne-Dorigny
Switzerland

Tel ++41 (0)21 692-3438
Fax ++41 (0)21 692-3305

Faculty page:
http://www.hec.unil.ch/people/jantonakis

Personal page:
http://www.hec.unil.ch/jantonakis
____________________________________________________



On 04.06.2010 14:58, Collewaert V (MCFE) wrote:
Dear Statalist,

I am testing a model based on team observations. However, teams consist of two clearly defined parties, similar to supervisors and subordinates. One of my reviewers suggested - instead of running a multilevel model on all observations at once - running seemingly related regression, using two equations, one for each 'subsample'. Sample sizes are not the same. In other words, the equation would be the same (same independent and dependent variables) but the observations based on which it would be estimated would be different, but related as they belong to the same overarching team. As I'm new to SUR, can anyone help me with the code I should be using for this? Should I use one stata file or two? Further, is there a way to control for non-independent observations within each equation? (as I sometimes have multiple subordinates and hence observations per team)
Any help would be greatly appreciated.

Kind regards,

Veroniek

*
*   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/
*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index