Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

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

Re: st: post estimation and standard errors

From   Maarten buis <>
Subject   Re: st: post estimation and standard errors
Date   Fri, 11 Jun 2010 14:43:27 +0000 (GMT)

--- On Fri, 11/6/10, Stefan wrote:
> Maybe someone can help. I want to meddle with my standard
> errors after regress to see how this affects the Wald test.
> Since Stata does not let me simply change the system
> variables _se[VAR] and then run the test and the Wald test
> takes only
> standard adjustments for the SEs as an argument I see no
> option besides coding or doing it manually. Is there a nicer
> way?

the graphs below show the consequences of changes in the 
standard error for the Wald statistic and resulting the 
p-value, assuming that the null hypothesis is that the
coefficient equals 0.

*------------------ begin example ----------------
sysuse auto, clear
logit foreign mpg rep78
twoway function y = _b[mpg]/x, ///
       range(.05 .5)           ///
       ytitle(Wald statistic)  ///
       xtitle(standard error)  ///

twoway function y = 2*normal(-abs(_b[mpg]/x)), ///
       range(.05 .5)                           ///
       ytitle(p-value)                         ///
       xtitle(standard error)                  ///
*------------------ end example --------------------
(For more on examples I sent to the Statalist see: )

Hope this helps,

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


*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index