PSUPLD is the pooled variance primary sampling unit variable that represents the impact of the sample design clustering on the estimates of variance and standard errors. It is appropriate to use PSUPLD instead of the annual primary sampling unit (PSUANN) if the years included in a pooled analysis include survey years 1996-2001 or 2019-forward. When pooling data from only 2002-2018, it is appropriate to use the variable PSUANN, the annual stratum variable for each survey year.
For analysis, researchers need to use PSUPLD in conjunction with STRATAPLD to account for stratification and clustering when computing variance estimates with IPUMS MEPS data.
PSUPLD should be used when analysts wish to pool together (aka combine) more than one year of MEPS annual data including any years before 2002 or after 2018. In most years, MEPS samples are not completely independent from one another, both because each panel is in the MEPS for two consecutive years and because households are sampled from the same geographic areas in multiple years. However, because MEPS is designed to be nationally representative in every year, multiple sample years can still be pooled. The use of PSUPLD and STRAPLD is necessary to obtain appropriate standard errors by specifying a common variance structure.
PSUPLD is a 1-digit numeric variable.
0: Not in Universe
- 1996-2019: All persons.
- 1996-2020: All persons.
- No weights are available for this variable. Please refer to PSUPLD for more information.