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 on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: Survey Design Degrees of Freedom


From   Steve Samuels <sjsamuels@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Survey Design Degrees of Freedom
Date   Thu, 5 Aug 2010 10:02:29 -0400

Unfortunately,  -svyset- will accept -dof()- without a bootstrap in
sight, and the -svy- estimation results will use the assigned degrees
of freedom, even if the value is greater than the number of
observations.  I suggest that StataCorp  rethink this option.

Steve



On Wed, Aug 4, 2010 at 4:20 PM, Stas Kolenikov <skolenik@gmail.com> wrote:
> As far as I understand, -dof()- only applies to -svy bootstrap-, since
> in that case the design d.f. is unknown. With specified PSUs, the
> design d.f. will be derived from the count of PSUs and strata.
>
> Korn & Graubard give some suggestions in their 1999 book. Not that I
> like these suggestions, but for the record, I have to give this
> reference. Looks like you know all this stuff already though.
>
> On Wed, Aug 4, 2010 at 11:43 AM, SEVIGNY, ERIC <SEVIGNY@mailbox.sc.edu> wrote:
>> Hello All-
>>
>> I am running an -mlogit- model on complex survey data. In estimating the
>> model, I am running into the all-too-common insufficient degrees of
>> freedom problem (I have 40 PSUs and 10 strata and thus 30 df).
>>
>> There have been a number of suggestions in the literature to work around
>> this issue, e.g., ignore either the strata or PSU boundaries, randomly
>> split each PSU into two pseudo-PSUs, etc.
>>
>> My question revolves around Stata's recent but poorly documented -dof()-
>> option with -svyset-. Regardless of its merit, let's say I want to more
>> than double my df by creating pseudo-PSUs. I could do this manually and
>> randomly divide the PSUs, giving 80 - 10 = 70 df. My question is this:
>> Can I leave the PSUs as is and simply include -dof(70)- in my -svyset-
>> command and achieve in effect the same result?
>>
>> Regards,
>>
>> Eric L. Sevigny, Ph.D.
>> Assistant Professor
>> Department of Criminology and Criminal Justice
>> University of South Carolina
>> 1305 Greene St
>> Columbia, SC 29208
>> (803) 777-7043 (office)
>> sevigny@mailbox.sc.edu
>>
>>
>> *
>> *   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/
>>
>
>
>
> --
> 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/
>



-- 
Steven Samuels
sjsamuels@gmail.com
18 Cantine's Island
Saugerties NY 12477
USA
Voice: 845-246-0774
Fax:    206-202-4783

*
*   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