I’m using data from supplemental surveys in 1988 and 1989 to determine mortality. Would it be better to use the supplemental sampling weights or the mortality weights for those years? My exposure of interest was only recorded in those supplements. Thank you in advance for any help/suggestions.
To my knowledge, there are no adjusted supplemental weights for the NHIS linked mortality files with these earlier years of supplement data (there is a weight for the 1997-2018 sample adult specific questions, but that doesn’t apply to your years of interest). My best suggestion is to email the NCHS data linkage team for guidance on how to adjust the supplement weights for mortality linkage eligibility. I am sharing some resources from the NCHS data linkage team’s page on linking NCHS data to NDI mortality files: