Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Jerome Ette <jerome.ette@googlemail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: ivreg vs ivregress |
Date | Wed, 1 Dec 2010 17:38:20 +0100 |
i always used the command ivreg for 2SLS analysis. now i recognized that it is termed 'out-of-date' in the current version and it is recommended to use the function ivregress instead. when i switch to ivregress i get different inference statistics than i used to get with ivreg. but of course the first stage regressions are the same, so are the residuals of both approaches and further post estimation tests. here is an replication example from the STATA help files. webuse hsng2 ivregress 2sls rent pcturban (hsngval = faminc) ivreg2 rent pcturban (hsngval = faminc) ivreg rent pcturban (hsngval = faminc) so why aren't the output tables for all three commands (ivregress, ivreg2 and ivreg) exactly the same? what has changed from ivreg to ivregress? or am i missing an (optional) argument in any of the commands? thanks! best jerome * * 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/