NHIS: pooling, complex sampling and weights


#1

I pooled the NHIS data from 2010-2017. Do I need to account for different design strategies (2010-2015; 2016-2017) for the PSU and STRATA variables when creating my complex design plan? Or has it already been taken care of? If IPUMS has already adjusted the PSU and STRATA variables, can I just divide weight variables (i.e. PERWEIGHT) by 8 to account for all 8 years?


#2

You only need to create a weight variable by dividing it by number of years (There is another more complicated way to create the weights). There is no need to change PSU and STRATA.