Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: lincom SEs getting mixed up in outreg


From   "Danielle H. Ferry" <dferry@nber.org>
To   StataList <statalist@hsphsun2.harvard.edu>
Subject   st: lincom SEs getting mixed up in outreg
Date   Thu, 24 Apr 2003 09:35:45 -0400

Dear Statalisters,

I'm executing a series of 6 tests (5 lincoms & 1 F-test) after a regress
statement. Then I ask outreg to report these results using the addstat
option. This is repeated 4 times (for 4 different specifications) so that my
resulting outreg output has 4 columns of results. Everything goes fine for
the first specification. In the other 3 specs, though, the SEs on the
lincoms are reported out of order. (Omitting coeff's) I get:

Col1    Col2    Col3    Col4
se1     se4     se2     se3
se2     se1     se1     se5
se3     se3     se3     se1
se4     se5     se5     se2
se5     se2     se4     se4

Previously, I had a problem where an entire lincom statement would repeat in
my outreg and Patrick Joly solved it. (It was due to the fact that outreg
was written for Stata 6 and only understands variable names of 7 characters
or less.) Perhaps this is related.

I'd really appreciate any help. This seems v. strange. Note that if I trim
the text in the addstat option (i.e. Lincom test 1, est1, SE, se1) - see
below - the problem remains but the order changes.

This is my outreg statement:

outreg2 _IpreCap_1 postCap
   using famCap4_parity1_race`i',
   title(Relative Changes in the Abortion Ratio Among Race=`i')
   ctitle(Base)
   addstat(Lincom test: postCap + parity>=1 + (parity>=1*postCap),
            est1, SE, se1,
           Lincom test: preCap + parity>=1 + (parity>=1*preCap),
             est2, SE, se2,
           Lincom test: (parity>=1*postCap) - (parity>=1*preCap),
             est3, SE, se3,
           Lincom test: (parity>=1*postCap) - (parity>=1*preCap) -
             parity>=1, est4, SE, se4,
           Lincom test: preCap + (parity>=1*preCap), est5, SE, se5,
           F-test: preCap + (parity>=1*preCap) = 0 & preCap = 0,
            fstat1, Prob, prob1)
   addnote("", "Base regression: ln of abortion ratio on age, parity,
           pre-cap,",
          "post-cap, unemployment rate, max AFDC benefit, year",
            "Alt1: base regression w/o time dummies",
            "Alt2: base regression w/ interaction of parity & year",
            "Alt3: base regression excluding NYC",
            "Run at $S_TIME, $S_DATE", Using data from $S_FN)
   br nocons se rdec(3) adec(3) replace;

-- 
Danielle H. Ferry
National Bureau of Economic Research
365 Fifth Avenue, 5th Floor
New York, NY 10016



*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index