Statalist


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

st: Non-Linear 2SLS?


From   Christian Weiss <christian.weiss@nightberry.de>
To   statalist@hsphsun2.harvard.edu
Subject   st: Non-Linear 2SLS?
Date   Tue, 12 May 2009 20:40:54 +0200

Dear statalisters,

 I am currently trying to get to know to the 2SLS method.
Among many things I am wondering, if it is possible to include
non-linear items (e.g. squared or cubic) in a 2SLS model.

Let's take the following equation as an example:

ivregress 2sls tobinsq (cr5= netincome5yr  volatilityofsales
totalassets industry) rndtosales

Let's assume I would like to investige a curvilinear effect of cr5.
How could I manage to do that?
In a standard OLS regression I would create the variable cr5_sqr
containing the squared value of cr5 (gen cr5_sqr = cr5^2).
However, if I square the cr5 variable in the 2SLS I am probably
messing with the effect of the instruments.

Any suggestions?
All the best
Christian
*
*   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   |   What's new   |   Site index