Hi,
I am using for an analysus all the cycles of the Tobacco Supplement (TUS) from 1995 to 2015. It is said that for each three month cycle of the TUS, in order to have a representative sample, the weights need to be divided by three, Which weight should I use for January, 2000; May, 2000 and May, 2011 if I use only the individuals that have not been interviewed in the TUS before (and not the longitudinal cohort)?
Thank you for your help,
Camille