Statalist The Stata Listserver

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

st: AW: multivariate linear reg

From   "Stephan Brunow" <>
To   <>
Subject   st: AW: multivariate linear reg
Date   Wed, 17 May 2006 08:17:09 +0200


if your model is

y1=a1 + b1*x + u
y2=a2 + b2*x + u
y3=a3 + b3*x + u

and x is the same for all equations you may estimate it equation by equation
OLS (regress y1 x1 and so on) or you may run sureg. (Seemingly unrelated
regression model). In both cases, as long as x is the same for all
equations, you should get the same results. Check Greene, Econometric
Analysis, Ch. 14. for details.
If you have various x in the different equations you should apply sureg. As
long as the error terms are uncorrelated, you may also apply OLS. However,
this is written in detail in the Greene-Text-Book.


Stephan Brunow
MSc. in Economics und Diplom-Verkehrswirtschaftler 
Professur für VWL, insb. Makroökonomik und
Institut für Wirtschaft und Verkehr
Fakultät für Verkehrswissenschaften „Friedrich List"
Technische Universität Dresden
D-01062 Dresden 

Phone: ++49-(0)351-463-36806
Fax: ++49-(0)351-463-36819

-----Ursprüngliche Nachricht-----
[] Im Auftrag von stata_user
Gesendet: Dienstag, 16. Mai 2006 19:36
Betreff: st:multivariate linear reg

I would like to ask about the difference between a simple linear regression 
of 3 depdent variables Y1 and Y2 and Y3 with a independent variable X

and the the multivariate version which regress (Y1,Y2,Y3) with X  (test 
statistics used in each case, interpretation of the output , if possible), 
and how to do it in stata?

many thanks in advance?

Are you using the latest version of MSN Messenger? Download MSN Messenger 
7.5 today!

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index