Status:Closed    Asked:Sep 21, 2016 - 02:51 PM

What is the correct modification to wtsupp in pooled analyses that restrict to mish==1,2,3,4

Due to the rotating design, a staff member suggested to limit analyses of pooled ASEC data to households in their first rotation (mish=1-4) to deal with duplicate cases. In an unrelated strand, people argue that the weights should be divided by the number of pooled years, but these use the entire cross-section (year) sample. So, what is the proper modification to the weights for analyses with pooled data (say 2005-2015) which restricts the analyses to households in the first rotation (mish=1,2,3,4). Do we multiply the weights times 2 to account for the fact that we have eliminated half the sample? Is it (wtsupp/12)*2?

Do you have the same question? Follow this Question

Staff Answer


Jeff Bloem


Your intuition seems sound. The WTSUPP weight variable can be divided by the number of years in the analysis and then, because restricting the data to households in the first rotation the data is dropping half the population, the weight can be multiplied by 2. Making these adjustments to the weight variable will not influence any regression results you may preform with this data. It will, however, provide you with a nationally representative estimate of specific variables within your sample.


Sep 22, 2016 - 10:26 AM

Report it


I am looking for the Kessler 6 variable in IPUMS-MEPS
Reason for missing MIGSTA1 values in ASEC files for 1985 (and 1995)
India Area-level data (geo2) inconsistent for total population, urban share, ...
Does anyone have a method for handling missing data in CPS "EDUC" variable? 2...
How do I find the data file:Ipumsi_00001.dat for the 1974
Variables from Full Year Consolidated Files currently not included in IPUMS-M...
Login   |   Register

Recently Active Members

View More »

Share |