st: Adjusting for individual clustering in pooled -svyset- data

[htunsta2@staffmail.ed.ac.uk]

Dear Statalist,

Id really appreciate advice please regarding the best approach to 
adjusting for individual clustering in a pooled cross-sectional -svyset- 
dataset.

Ive pooled data from 10 waves of the British Household Panel Survey 
(BHPS) (waves 6-15). The pooled survey waves that have been combined 
therefore are not independent and each individual BHPS respondent 
appears 1-10 times in the pooled data.

BHPS has a complex survey design, so I’ve -svyset- the data and am using 
the relevant individual level cross-sectional weights for each wave.

Im completing -svyset- logistic regression analysis (-logistic-).

The analysis is being carried out in Stata 12.1 (64 bit).

Im using the Stata commands below to -svyset- the data:
. svyset [pweight = weightvar], psu(psuvar) strata(stratavar) 
singleunit(scaled)

Im uncertain though about the best approach to adjusting for 
individuals’ repeated appearances in the dataset when the dataset is 
also being adjusted for geographical clustering resulting from the 
complex sample design. What Stata code should I be using to take account 
of individual clustering in the -svyset- data?

Thank you very much,
Helena Tunstall

--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

*
*   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/