Hello, I downloaded data from IHIS from 2000-2010 and noticed the PSUs are limited to 2 values per year. Is this accurate?
Thanks,
Brent
Yes, this is correct. The PSU and STRATA variables, used in conjunction, are useful for accounting for stratification and clustering when computing standard errors associated with estimates from IPUMS NHIS data. More detail about variance estimation with PSU and STRATA is available on this page.
Thanks Jeff, Do you know why there are only 2 PSUs per year from 2000 to 2010?
Since 1995, the NHIS sampling frame has divided the US into strata roughly corresponding to each of the 50 states and Washington, DC. Primary sampling units are nested within strata, with no more than two PSUs per stratum. When generating population estimates based on the NHIS data, you should specify both STRATA and PSU to account for the complex design of the NHIS. For more information on the NHIS sample design covering the 2000-2010 period, you can refer to the following NCHS methodological papers:
Botman SL, Moore TF, Moriarity CL, and Parsons VL. Design and Estimation for the National Health Interview Survey, 1995-2004. National Center for Health Statistics. Vital Health Stat 2(130). 2000. https://www.cdc.gov/nchs/data/series/sr_02/sr02_130.pdf
Parsons VL, Moriarity C, Jonas K, et al. Design and Estimation for the National Health Interview Survey, 2006-2015. National Center for Health Statistics. Vital Health Stat 2(165). 2014. https://www.cdc.gov/nchs/data/series/sr_02/sr02_165.pdf