Re: st: VEC: t- instead of z-statistics

[Barbara Engels <engels.ba@gmail.com>]

Thanks John, thanks Richard. Richard, this exactly is my concern. On top, I have only a relatively small number of observations ... In my opinion, there SHOULD be a way to demand t-statistics.

El 22.07.2011, a las 06:49, Richard Williams escribió:

> At 10:27 PM 7/21/2011, John Luke Gallup wrote:
>> The coefficient test statistics reported after most estimation models are z-statistics because we only know their asymptotic distribution (even if we assume that the errors are normally distributed).  Only with simple linear models, like OLS, where the coefficients are linear functions of the y's, and only when we assume the errors are normally distributed, do the coefficient test statistics have a known small sample distribution - a t-distribution.  Note that a t-statistic is equal to a z-statistic asymptotically.
>> 
>> Since the VEC is not a linear estimator, it does not report t-statistics.
>> 
>> Please join in if you see any problems with what I said.
> 
> Potential gap/flaw in that reasoning? When data are svyset, we get t-statistics rather than z-stats, e.g.
> 
> webuse nhanes2f
> svy: logit diabetes age black female
> 
> I'm not sure what the moral of that is, but Stata does report t statistics in many more situations than just after OLS.
> 
> 
> -------------------------------------------
> 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/