Bookmark and Share

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


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

Re: st: Making vce(robust) the default


From   Billy Schwartz <[email protected]>
To   [email protected]
Subject   Re: st: Making vce(robust) the default
Date   Thu, 4 Aug 2011 13:41:56 -0400

Mr. Kolenikov's suggestion is consistent with how StataCorp wrote the
regress command. Typically when I get stumped about how to write code,
I just read their ado files. Type

viewsource regress.ado

and look at the syntax command or just search (ctrl-F) for "vce" or "robust"

On Thu, Aug 4, 2011 at 12:38 PM, Stas Kolenikov <[email protected]> wrote:
> I guess that if you are concerned with a particular standard option of
> Stata, you'd still want to parse the inputs from the user yourself:
>
> syntax ... , [vce(passthru) robust cluster(passthru)]
>
> * with appropriate capitalization per Stata conventions
>
> if "`vce'`robust'`cluster'" == "" local vce vce(robust)
>
> ml ... , ... maximize `vce' `robust' `cluster'
>
> This should digest the user-provided -vce- option and its variants.
> You'd have to ask Jeff P if there is any way to figure out whether a
> program is called by -svy-. As long as this is a very deep
> technicality that is dependent on the particular
> implementation/version of -svy-, subject to change without notice,
> there may no be such an official way. My understanding though is that
> for the linearization estimator, -svy- is going to call -predict,
> score- when computing the standard errors, so it may not matter that
> much whether you enforce -robust- for maximization purposes.
>
> On Thu, Aug 4, 2011 at 12:07 PM, Richard Williams
> <[email protected]> wrote:
>> At 10:53 AM 8/4/2011, Billy Schwartz wrote:
>>>
>>> Use the NOoption form (help syntax##optionally_off). Call the option
>>> "norobust" and write
>>>
>>> if "`robust'" == "" local robust vce(robust)
>>
>> Thanks. But suppose somebody had specified something like -vce(cluster
>> clustervar)-? Or had explicitly specified -vce(oim)-? I don't want to zap
>> anything the user has implicitly or explicitly specified. Rather, in the
>> absence of any implicit or explicit declaration of vce, I want the default
>> to be robust, not oim.
>>
>
>
>
> --
> Stas Kolenikov, also found at http://stas.kolenikov.name
> Small print: I use this email account for mailing lists only.
> *
> *   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 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index