Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


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

Re: st: margeff and margins puzzle


From   Patrick Roland <patrick.rolande@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: margeff and margins puzzle
Date   Fri, 29 Jul 2011 09:36:50 -0700

This is very odd. I am using Stata 11.2 and my version of margeff is up to date:

. which margeff
*! Obtain partial effects after estimation
*! Version 2.2.0  (20 August 2009)    (Revision of Stata Journal submission)
*! Author:        Tamas Bartus        (Corvinus University, Budapest)

The results I get from margins are the same as you, Richard. The
results I get from margeff are:

. margeff

Average partial effects after probit
      y  = Pr(y)

----------------------------------------------------------------------------------------------------
    variable |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
          x1 |          0   .0518072     0.00   1.000    -.1015403    .1015403
----------------------------------------------------------------------------------------------------




On Thu, Jul 28, 2011 at 7:11 PM, Richard Williams
<richardwilliams.ndu@gmail.com> wrote:
> At 06:35 PM 7/28/2011, Patrick Roland wrote:
>>
>> Hi all,
>>
>> Here's a piece of code which gives peculiar results. It's a simple probit.
>> "margeff" calculates the marginal effect to be exactly zero, whereas
>> margins
>> (correctly) doesn't. The really puzzling thing is that if "set obs 1000"
>> is
>> changed to "set obs 100", they give identical results, as they should!
>> Anyone have any idea what might be going on here?
>>
>>
>> set seed 1
>> set obs 1000
>> gen x0 = -.25
>> gen x1 = runiform()
>> gen eps = rnormal()
>> gen y = x0+x1+eps > 0
>> probit y x1
>> margeff
>> margins, dydx(*)
>
> I get identical results with both commands. As always, make sure you have
> the latest versions of everything and try again.
>
> . set seed 1
>
> . set obs 1000
> obs was 0, now 1000
>
> . gen x0 = -.25
>
> . gen x1 = runiform()
>
> . gen eps = rnormal()
>
> . gen y = x0+x1+eps > 0
>
> . probit y x1
>
> Iteration 0:   log likelihood = -669.63431
> Iteration 1:   log likelihood = -637.00519
> Iteration 2:   log likelihood = -636.94994
> Iteration 3:   log likelihood = -636.94994
>
> Probit regression                                 Number of obs   =
> 1000
>                                                  LR chi2(1)      =
>  65.37
>                                                  Prob > chi2     =
> 0.0000
> Log likelihood = -636.94994                       Pseudo R2       =
> 0.0488
>
> ------------------------------------------------------------------------------
>           y |      Coef.   Std. Err.      z    P>|z|     [95% Conf.
> Interval]
> -------------+----------------------------------------------------------------
>          x1 |   1.133848   .1423649     7.96   0.000     .8548182
>  1.412878
>       _cons |  -.2788769   .0797004    -3.50   0.000    -.4350869
>  -.122667
> ------------------------------------------------------------------------------
>
> . margeff
>
> Average partial effects after probit
>      y  = Pr(y)
>
> ------------------------------------------------------------------------------
>    variable |      Coef.   Std. Err.      z    P>|z|     [95% Conf.
> Interval]
> -------------+----------------------------------------------------------------
>          x1 |   .4126165   .0466284     8.85   0.000     .3212265
>  .5040065
> ------------------------------------------------------------------------------
>
> .
> . margins, dydx(*)
>
> Average marginal effects                          Number of obs   =
> 1000
> Model VCE    : OIM
>
> Expression   : Pr(y), predict()
> dy/dx w.r.t. : x1
>
> ------------------------------------------------------------------------------
>             |            Delta-method
>             |      dy/dx   Std. Err.      z    P>|z|     [95% Conf.
> Interval]
> -------------+----------------------------------------------------------------
>          x1 |   .4126127   .0466285     8.85   0.000     .3212225
>  .5040028
> ------------------------------------------------------------------------------
>
>
>
> -------------------------------------------
> Richard Williams, Notre Dame Dept of Sociology
> OFFICE: (574)631-6668, (574)631-6463
> HOME:   (574)289-5227
> EMAIL:  Richard.A.Williams.5@ND.Edu
> WWW:    http://www.nd.edu/~rwilliam
>
> *
> *   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/


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