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: Survey standard errors of grouped data proportions


From   Philip Burgess <filip.liljeholm@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Survey standard errors of grouped data proportions
Date   Tue, 17 Dec 2013 08:49:36 +1000

Thanks Stas;

Thought it was simple.

Much appreciated.

Philip


On Tue, Dec 17, 2013 at 12:06 AM, Stas Kolenikov <skolenik@gmail.com> wrote:
> After -svy: total-, you can use -lincom- and -nlcom- to get the right
> statistics and the right standard errors:
>
> svy : total  weight , over(region, nolab)
> local total (_b[weight:1] + _b[weight:2] + _b[weight:3] + _b[weight:4])
> lincom `total'
> nlcom _b[weight:1]/ `total'
>
>
> -- 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, Dec 16, 2013 at 1:57 AM, Philip Burgess
> <filip.liljeholm@gmail.com> wrote:
>> I am using Stata 13.1 SE on a Windows 7 platform.
>>
>>
>> I have complex survey design data set, with replicate weights, etc and have
>> successfully svyset using the statement
>>
>>
>> svyset   _n [pweight=3Dwgt], vce(linearized) singleunit(missing)
>>
>>
>> I have two variables of interest, patient visits (continuous) and patient
>> severity group (5 mutually exclusive categories).
>>
>>
>> To estimate the average number of patient visits, I simply run:
>>
>> svy: mean visits, over(group)
>>
>> This gives weighted means and appropriate standard errors (and thus the CI)
>>
>>
>>
>> What I next want to estimate in the relative proportion of all patient
>> visits in each of the five groups.
>>
>> For example, the most severe patient group accounts for approximately 60%
>> of all patient visits.
>>
>> I can estimate the overall total of patient visits per group using the
>> command:
>>
>> svy: total visits, over(group)
>>
>> and from that can work out the point estimate of the relative proportion
>> for each of the 5 groups.
>>
>> But I cannot figure out how to get the SE of the relative proportion.
>>
>>
>> This is probably incredibly simple but I have had no luck searching the
>> FAQ, the manual or the list archives.
>>
>> Thanks;
>>
>> Philip
>> *
>> *   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/



-- 
+++++++++++++++

Philip Burgess
*
*   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/


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