Re: st: Issue with 'bsweights'

[Stas Kolenikov <skolenik@gmail.com>]

You're welcome. I typically use -bsweights- with -balanced- option as
it runs faster that way (fully programmed in Mata, as opposed to
Stata's -bsample, weight-) and provides minor improvements in biases
of the estimators.

-- Stas Kolenikov, PhD, PStat (ASA, SSC)
-- Senior Survey Statistician, Abt SRBI
-- Opinions stated in this email are mine only, and do not reflect the
position of my employer
-- http://stas.kolenikov.name



On Tue, Nov 26, 2013 at 1:53 AM, Owen Gallupe <ogallupe@gmail.com> wrote:
> Oh! I see now. Somehow I had convinced myself (even after reading your
> article) that the 'nosvy' option wasn't appropriate in that situation.
> I now run the following and it appears to work nicely.
>
> bsweights bsw, reps(2000) n(-1) seed(22222) nosvy
> svyset sch [pw=wgt], vce(bootstrap) bsrweight(bsw*)
> svy: reg dv var1 var2 var3
>
> Thanks!
>
> Owen
>
>
> On Mon, Nov 25, 2013 at 10:59 AM, Stas Kolenikov <skolenik@gmail.com> wrote:
>> -bsweights- is intended to work with the survey data first and
>> foremost. As I describe in the Stata Journal article supporting the
>> package, you can specify -nosvy- to run -bsweights- without trying to
>> pull out the -svy- settings, and treating the data as i.i.d. The
>> -webuse nhanes2- data comes with -svy- settings glued into the data
>> set, and -bsweights- is able to pick that information up.
>>
>> -- Stas Kolenikov, PhD, PStat (ASA, SSC)
>> -- Senior Survey Statistician, Abt SRBI
>> -- Opinions stated in this email are mine only, and do not reflect the
>> position of my employer
>> -- http://stas.kolenikov.name
>>
>>
>>
>> On Mon, Nov 25, 2013 at 9:54 AM, Owen Gallupe <ogallupe@gmail.com> wrote:
>>> Please let me know if there is any other information that might help
>>> in coming up with a solution.
>>>
>>> Thanks.
>>>
>>> Owen
>>>
>>>
>>> On Mon, Nov 25, 2013 at 2:36 AM, Owen Gallupe <ogallupe@gmail.com> wrote:
>>>> Hi,
>>>>
>>>> I am having an issue using the 'bsweights' command that has me puzzled
>>>> given the documentation I've read and my own testing. When I try to
>>>> run the following with my data,
>>>>
>>>> bsweights bsw, reps(2000) n(-1) seed(22222)
>>>>
>>>> I get an error message:
>>>>
>>>> svy is not set
>>>> r(119);
>>>>
>>>> So I set up svyset and then run bsweights again but with the same error message.
>>>>
>>>> When I try the exact same command with the nhanes2 data set, it works fine.
>>>>
>>>> If anyone could set me straight, I would really appreciate it.
>>>>
>>>> Regards,
>>>>
>>>> Owen Gallupe
>>>>
>>>> p.s., I'm running Stata/SE 12.1 on Windows 7 (64 bit)
>>>> *
>>>> *   For searches and help try:
>>>> *   http://www.stata.com/help.cgi?search
>>>> *   http://www.stata.com/support/faqs/resources/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/faqs/resources/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/faqs/resources/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/faqs/resources/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/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/