I am analyzing a subpopulation of the 2014 ACS 1% sample, and I would like to use replicate weights to calculate my standard errors. Can I still use replicate weights after using the Select Cases feature on IPUMS-USA, or do I have to download the entire ACS, and then use Stata’s svy, subpop() option?
As mentioned in the Replicate Weight FAQ page, accurate calculation of standard errors using replicate weights requires all records from the sample to be present. So you will need to include all cases from a sample to use replicate weights. In my own testing I have generated identical standard errors using a full sample and a sample using selected cases, however it is always recommended to use the full sample as Stata’s BRR methods check that there are two PSUs in every stratum (this is built into the replicate weights provided by the Census Bureau) and it is possible to create errors when this is not the case in a particular subset of respondents.
I hope this helps.