Hi. I'm preparing to run a probit regression in Stata using ACS PUMS data.

Does anyone have experience using the Stata svyset command (or, more generally, specifying relevant survey design factors in a statistical analysis program) with PUMS files?

The following is suggested for Current Population Survey analyses (see https://www.stata.com/statalist/archive/2008-04/msg00444.html), which uses state (gestcen) and consolidated statistical area (gtcsa) variables.

egen psu=group(gestcen gtcsa) svyset [pw=mars], strat(gestcen) psu(psu)

Does anyone have an analogous specification for PUMS?