Sample weighting

#1

I’m pooling 2011-2015 data from IPUMS IHS. I also read other people’s posting on this particular issue. However, I also again would like to ask a sample weighting, while running analyses. My syntax is as follows:
svyset psu [pweight=pooledsampweight], strata(strata)
where pooledsampweight is created by the following ways:
gen pooledsampweight= sampweight/5
Would you please mind whether I am doing what it should be?
Thank you very much.
Hari Poudel

#2

Your approach here is certainly sensible. A more generalized way of defining what you a proposing is to multiply the cross-sectional sampling weight in sample i by (sample size in sample i)/(sample size of pooled sample). Since the NHIS samples are all roughly the same size, this approach will be quite close to what you have proposed above. More detail on this topic can be found on this page under the “Pooling Multiple Years of Data” section.

1 Like
#3

Thank you Jeff!