Since the dataset I'm using has a Stratum (v022) with only one PSU (v021), and I have to use the following four methods to solve the problem of missing standard error and test statistic.
1. svyset [pw=wgt], psu( v021) strata( v022) singleunit(centered)
2. svyset [pw=wgt], psu( v021) strata( v022) singleunit(scaled)
3. svyset [pw=wgt], psu( v021) strata( v022) singleunit(certainty)
4. By reassigning the stratum with single PSU to another nearby stratum
My query is which one of the above methods (centered, scaled, certainty and reassigning stratum) is the most preferred that will give an efficient results.
1. svyset [pw=wgt], psu( v021) strata( v022) singleunit(centered)
2. svyset [pw=wgt], psu( v021) strata( v022) singleunit(scaled)
3. svyset [pw=wgt], psu( v021) strata( v022) singleunit(certainty)
4. By reassigning the stratum with single PSU to another nearby stratum
My query is which one of the above methods (centered, scaled, certainty and reassigning stratum) is the most preferred that will give an efficient results.
