I am planning to use data from multiple waves of data collected by the Add Health study. If I specify this as the sample weight in the survey design, I believe R studio will only use data from respondents with this weight. However, we are also supposed to specify that we are using data that is effectively a subpopulation of the total sample (for accurate estimates of standard error).
Does specifying a subpop make sense if the design leads to the selection of participants with a survey weight which is basically creating a subpop since only a fraction of participants have the weight.
Does this make sense? Hopefully I am understanding and explaining this accurately.
Thank you for any assistance you can offer.