Dear Statalists,
I am running analysis on a survey dataset which use multi-stage stratified systematic sampling design, and stratified sampling with Probability Proportional to size(PPS) depending on county population(first stratum).
There are 3 strata in my dataset, which are county, strata, and psu_id. Strata (second stratum) is nest within county (first stratum). This survey use stratified random sampling from strata (second stratum), and pull out third stratum named psu_id, and the number of each sampling from second stratum are named psu.

The stratum message are as follow:
First stratum: county (number 1 to 23)
Second stratum: strata (number 1 to 3)
Third stratum: psu_id
psu: the number of each sampling from second stratum

if the same number of both county and city, psu is the same. However, if the same number of only county or city, psu is different.

I use stata code as follows:
svyset county[pweight=WT_N],strata(strata) fpc(psu)||psu_id
However, the result will come out with “ fpc for all observations within a stratum must be the same”

How can I handle the problem with fpc is different. Could stata use in survey sampling with PPS design?

I look forward to hearing from you.
Thanks.

Lily