Hi, and welcome!
Please see the FAQ: What's a reproducible example (`reprex`) and how do I do one? Using a reprex, complete with representative data will attract quicker and more answers.
It's difficult to answer this in the abstract, because it depends on the data so much.
If you have geocoding of survey respondents in the US you can use tidycensus or another package to get population characteristics. For example, in some data I've worked with, the white, non-hispanic population serves as a useful marker. You would then be able to demographically weight as group/total. Also see the CRAN Task View for packages that may have useful tools for this.