RE: st: joint estimation of OLS and logit in seemingly unrelated regressions

[#WANG ZITIAN# <WANG0536@e.ntu.edu.sg>]

Thanks Maarten. 

However, I find that using suest command with clustering generates almost the same standard errors comparing estimating each individual regression with clustering option. Could you explain why it is the case? Or I made a mistake here ?

. cap drop _est*

. reg rnd etrm_before
------------------------------------------------------------------------------
         rnd |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
etrm_befor~y |   .1490155   .0154133     9.67   0.000     .1187669    .1792641
       _cons |    .040344    .003745    10.77   0.000     .0329945    .0476936
------------------------------------------------------------------------------

. est store ols

. logit apply  etrm_before_apply

------------------------------------------------------------------------------
       apply |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
etrm_befor~y |   4.738024   .6010609     7.88   0.000     3.559966    5.916082
       _cons |  -.8450138   .1036933    -8.15   0.000    -1.048249   -.6417786
------------------------------------------------------------------------------

. est store logit

. suest ols logit, vce (cluster stkcd_order)

Simultaneous results for ols, logit

                                                  Number of obs   =        939

                          (Std. Err. adjusted for 672 clusters in stkcd_order)
------------------------------------------------------------------------------
             |               Robust
             |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
ols_mean     |
etrm_befor~y |   .1490155   .0280263     5.32   0.000     .0940849    .2039461
       _cons |    .040344   .0041447     9.73   0.000     .0322205    .0484675
-------------+----------------------------------------------------------------
ols_lnvar    |
       _cons |  -4.860371   .1396304   -34.81   0.000    -5.134041     -4.5867
-------------+----------------------------------------------------------------
logit_apply  |
etrm_befor~y |   4.738024   .8938238     5.30   0.000     2.986161    6.489886
       _cons |  -.8450138    .128236    -6.59   0.000    -1.096352   -.5936759
------------------------------------------------------------------------------

. 
end of do-file

. logit apply  etrm_before_apply, cluster(stkcd_order)

                          (Std. Err. adjusted for 672 clusters in stkcd_order)
------------------------------------------------------------------------------
             |               Robust
       apply |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
etrm_befor~y |   4.738024   .8938238     5.30   0.000     2.986161    6.489886
       _cons |  -.8450138    .128236    -6.59   0.000    -1.096352   -.5936759
------------------------------------------------------------------------------

. reg rnd etrm_before, cluster(stkcd_order)

                          (Std. Err. adjusted for 672 clusters in stkcd_order)
------------------------------------------------------------------------------
             |               Robust
         rnd |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
etrm_befor~y |   .1490155   .0280413     5.31   0.000     .0939563    .2040747
       _cons |    .040344   .0041469     9.73   0.000     .0322015    .0484865
------------------------------------------------------------------------------

-----邮件原件-----
发件人: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] 代表 Maarten Buis
发送时间: 23 May 2012 20:21
收件人: statalist@hsphsun2.harvard.edu
主题: Re: st: joint estimation of OLS and logit in seemingly unrelated regressions

On Wed, May 23, 2012 at 2:10 PM, #WANG ZITIAN# wrote:
> I want to apply seemingly unrelated regression method to two regressions, one being OLS and another one being logistic regression.  I don't think the "sureg" can do this as "sureg" is designed for two OLS regressions.  Is there a code/command in Stata that can deal with my case?

-help suest-

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany


http://www.maartenbuis.nl
--------------------------

*
*   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/